09-10-2024, 06:28 AM
(This post was last modified: 09-23-2024, 05:12 PM by Bazooka_Joe. Edited 2 times in total.)
Disclaimer: I am not a statistician, this study is meant to be interesting and somewhat entertaining, but the results are not intended to be taken as scientific fact. Use the information at your own peril as I will not accept any responsibility for the lack of success of any rosters constructed using the conclusions found in this study, however, I will happily accept full responsibility for any success that comes from it.
Any sleights, perceived or otherwise, against specific teams or player groups are not intended to offend and are merely either what the data shows (data doesn’t lie) or ill-conceived attempts at humour (I blame the culture barrier).
Introduction
So, I guess the first question I need to answer is: why? Why have I decided to scour the rosters of the past 6 seasons for TPE totals? The truth of the matter is, I was starting a different media project that required data on teams TPE allocation to different position groups. After being bounced around various senior and long-standing league members, I unfortunately came to the conclusion that the data I was after did not exist. This left me with 2 choices; either I give up on the media project as a whole, or I sack it up and collect/process the data myself. At this point it was suggested to me that I could automate this process, but I unfortunately have to admit that my computer skills shame my generation and are more in line with those expected of a 60 year old, and are certainly not able to complete such a task.
Fortunately for me, I am used to brainlessly and stubbornly brute forcing my way through large amounts of data for league media, having produced a number of statistically centred media pieces early in my league career.
I started by using the ISFL index to record the rosters for all ISFL teams over the last 6 seasons: S44, S45, S46, S47, S48 and S49. I then used the schedules from each season to determine the latest date a player could be updated before week 1 of each season. I'm aware that players can earn TPE over the season so this may not be the most accurate representation of TPE, but I felt it made sense as the baseline for a rosters TPE as otherwise I felt like it would become too complicated to keep track of. The dates for prior to week 1 updates were as follows:
S44 - Oct 30th
S45 - Nov 25th
S46 - Jan 27th
S47 - Mar 23rd
S48 - May 18th
S49 - Jul 13th
Once I had a comprehensive list of players for each season, I was able to cross reference this with the above dates in the TPE tracker, in order to record each player's TPE on week 1 of each listed season. Once I had completed this, I had a complete roster with TPE for each team over the last 6 seasons, accurate as of week 1 of each relevant season.
Yes, this was time consuming, it has cost me my social life, distracted me at work and put a strain on a number of relationships over the last few weeks. When I close my eyes I see depth charts, I have TPE numbers and week 1 dates keeping me awake at night, I have player career paths embedded into my brain that went IA 3 seasons ago - I may need help.
Ultimately, this then allowed me to get an overview of which teams had how much TPE allocated to each position group over the last 6 seasons, and how this aligned with their success in the regular season, in the form of win totals. I decided to only use the regular season for ease of data extraction, and also because we all know that the playoffs are a crap shoot and the sim gods toy with us - I am not bitter at all, I promise.
For simplicity's sake, I have labelled all teams that were a part of the Chicago/Osaka franchise as OSK, regardless of whether they existed before or after the re-brand.
General Impressions
The first thing I did was produce ISFL TPE heat maps for each team in each of the last 6 seasons. I will attach these below, and am aiming to complete the same in the coming seasons so this isn't as big a task should anyone want to look into this kind of data again.
Once I had the TPE heat maps, and a master table, completed I was able to start analyzing the data. I started by combing through the raw data and pulling out any data points that stood out to me, before I then went on to process the data more formally. I am going to start by discussing the data points that stood out to me on first pass.
The TPE metrics I extracted from the data set where as follows:
Total TPE - sum of TPE of all players on a roster in a given season
OFF TPE - sum of TPE of all players on offense in a given season
DEF TPE - sum of TPE of all players on defense in a given season
ST TPE - sum of TPE of all players on special teams in a given season
QB TPE - sum of TPE of all players assigned QB in a given season
RB TPE - sum of TPE of all players assigned RB in a given season
WR TPE - sum of TPE of all players assigned WR in a given season
TE TPE - sum of TPE of all players assigned to TE in a given season
OL TPE - sum of TPE of all players assigned to OL in a given season (this includes T, G and C)
DL TPE - sum of TPE of all players assigned to DL in a given season (this includes both DE and DT)
LB TPE - sum of TPE of all players assigned to LB in a given season
DB TPE - sum of TPE of all players assigned to DB in a given season (this includes both CB and S)
>1000 TPE - number of players on a roster over 1000 TPE, regardless of position
Avg TPE - TOTAL TPE divided by total number of players on the roster in a given season
I only included TPE allocated to user controlled players, not bots. This included both active players and IA players, regardless of whether drafted or signed via FA.
The raw data can be found at the below links if anyone is interested:
Rosters:
https://docs.google.com/spreadsheets/d/1...drive_link
Master Table - S44-S49:
https://docs.google.com/spreadsheets/d/1...drive_link
Averages
The first piece of data I wanted to look at was an average, as I would therefore then have something to compare the other numbers to. The means are as follows:
- Total 16589
- OFF 6953
- DEF 9083
- QB 1064, RB 1687, WR 2344, TE 395, OL 1467
- DL 2163, LB 2912, DB 4009
- ST 548
- >1000 players 5, average TPE per ISFL player 716
The modal ranges are as follows:
- Total 15000 - 17500
- OFF 6000 - 8000
- DEF 8000 - 10000
- QB 1000 - 1200, RB 1500 - 2000, WR 2500 - 3000, TE 0, OL 1000 - 2000
- DL 2000 - 2500, LB 2500 - 3000, DB 3000 - 5000
- ST 250 - 500
- >1000 players, average TPE per ISFL player
Total TPE
I'm going to start by talking about the top of the dataset, and a few teams that stood out to me TPE-wise across this period:
NOLA:
One of the repeated names at the top of the rankings or Total TPE are the New Orleans Second Line, also referred to as NOLA. When looking at the last 6 seasons, a few periods of dominance stand out, and S47-49 NOLA is perhaps the best of the bunch. When including data from the past 6 seasons, NOLA take 3 of the top 4 spots for total TPE with their rosters in the 3 most recent seasons. This spell at the top has lead to impressive win totals, as well as back to back championships in the last 2 seasons (I know I said playoffs don't matter, I'm a hypocrite, get over it).
NOLA have been a bit of a Cinderella story over the time period that falls under the scope of this study, reaching the aforementioned spell near the top of the rankings at the end of the assessed period, but starting in S44 with a bottom 10 total TPE for this entire 6 season stretch. The NOLA roster in S44 is also the only roster in this dataset to fail to record a single player with 1000 or more TPE on week 1 of the season.
When going back through the index to construct the rosters, one thing in particular stood out to me about NOLA in the 6 season stretch - they have consistently had low roster turnover season to season, losing and replacing a maximum of 1 or 2 players a season. This has allowed the roster to earn and grow together, with the same nucleus of the team present in S44, when they had 0 players over 1000 TPE, right through to S48/49 when they had 11 players over 1000 TPE. While this approach to roster building has clearly worked, it does leave you wondering where they'll be when this nucleus enters regression. Could we see NOLA back at the bottom of the pile in the next few seasons?
Other notable periods of dominance over this time period include S45-47 OCO and S44-45 HON. These 3 teams (S47-49 NOLA, S45-47 OCO and S44-45 HON) make up the top 8 total TPE rosters over the last 6 seasons. These 8 teams won a total of 88 games, averaging 11 wins per season over this stretch. For those looking for an early answer to ‘does TPE matter’, the early evidence certainly looks promising.
A few things I noticed in common for these 8 teams:
- All have over 1000 TPE allocated at QB
- All have over 2500 TPE allocated to WRs
- 6/8 teams have more than 3000 TPE allocated to OL
- All 5 non-OCO teams all have ~3000+ TPE at DL
- General trend is less TPE at RB and DB - 4/8 teams have under league average TPE at RB
NOLA stand out amongst top 8 for having a lot of TPE at RB - only teams to have over 2000 TPE at RB are both NOLA (S48 and S49)
OCO:
In the early portion of this 6 season stretch, the Orange County Otters (also called OCO) were frequently at the top of the rankings, both in terms of TPE and record. In fact, S45, S46 and S47 OCO rosters are 3 of the 9 rosters to start a season with over 20000 Total TPE in this period. However, despite consistent appearances at the top of the total TPE lists across this period, OCO didn’t build their rosters the same as everyone else.
The first thing I noticed with OCO, especially with their successful teams, was a relative lack of TPE allocated to DL. Despite their success in these seasons, OCO were twice in the bottom 11 of DL TPE over this period - 1461 TPE in S45 and 1392 in S46. These numbers are both well below league average at DL, and make their successes that much more impressive. Their successes on defense in spite of this may be explained by the measures taken to mitigate this potential weakness. S46 OCO had the highest TPE at LB over this 6 season period at 4953, and OCO S45 had the highest TPE at DB over the last 6 seasons at 6275. Whether this was an intentional approach or not, making the rest of the roster on defense some of the strongest groups we have seen in the last 6 years certainly won’t have hurt their attempts to mitigate their weakness on the DL. Interestingly, the only other team to have more than 6000 TPE invested at DB in this period is S46 OCO.
It wasn’t just on defense where OCO went against the grain. S46 OCO were second in Total TPE over this period but are the only team in top 10 for total TPE in the past 6 seasons to be outside the top 50 for TPE at RB position. OCO noticeably invested TPE on the OL, with both S46 and S45 OCO recording over 3000 TPE to start the season (2 of 8 total teams) and S47 narrowly outside (2912), meaning OCO take one third of the top 9 spots for OL TPE in this period. Not many teams prioritised OL TPE, so this data point stood out to me.
HON:
The Honolulu Hahalua, also referred to as HON, were almost the reverse NOLA over this time period, starting off very strong in S44 and S45 and then undergoing a dramatic rebuild *cough* tank *cough* in S46. To illustrate just how dramatic this change was, S44 and S45 HON appear in the top 8 for total TPE. HON then fail to appear in the top 50 for Total TPE over the following FOUR seasons, achieving a high of 55 with S49. The other 3 HON rosters (S46, S47 and S48) are all in bottom 15 for total TPE over the assessed period.
An interesting data point from the Hahalua, S44 and S45 HON are only teams in top 11 total TPE to have 0 TPE allocated to the TE position (more on this later).
Moving on, unfortunately for some of you reading this, it is time to talk about the bottom of the pile:
COL:
The Colorado Yeti, or COL, have fallen on hard times recently, and it is probably not a surprise to anyone that has paid attention to rosters, win totals, or the draft over the last couple of seasons to see them at the bottom of the table when it comes to looking at TPE and wins.
COL have the 2 lowest, and 3 of the bottom 5, total TPE rosters in the past 6 seasons (S47, S48 and S49). S47 COL are the only team in the last 6 seasons to record less than 11000 Total TPE on a roster.
Some of you may have already noticed this but yes, this means that 13 teams defenses, and S48 NOLAs offense, had more TPE as a unit than S47 COLs whole roster (sorry COL, but this data point just cannot be ignored).
One thing that stood out to me when looking at COLs rosters was the notably low TPE at WR. S47 and S48 COL are the only teams in the last 6 seasons to have less than 1000 TPE total at WR.
The low TPE totals are reflected in the records of these teams, most notably as S47, S48 and S49 COL are the only teams to have won less than 2 games in a season over this period.
I do want to give a shoutout to COL GM Aeonsjenni for having the only 1000 TPE player on COL at the start of each of the past 3 seasons.
NYS:
The other team to join COL at the bottom of the rankings are NYS, who claim the other 2 spots in the bottom 5 for Total TPE over the past 6 seasons. Despite having posted 11 wins over these 2 seasons, S48 and S49 NYS are the 3rd and 4th lowest total TPE rosters of the past 6 seasons. This makes their win total that much more impressive, as NYS have no single green position group on the roster through S48 and S49 except for K/P.
One other thing I noticed when looking at the rosters at the bottom of the table for Total TPE was that the bottom 5 teams have notably very little TPE allocated to OL. The league average for TPE across the OL group as a whole is 1467, but the totals for the bottom 5 teams are 0, 261, 0, 171 and 228. I know the league generally struggles to find active earners on the OL, and it may just be that during a rebuild teams prefer to use their active players at positions they deem more valuable, but I thought it was an interesting data point regardless.
Total TPE vs wins - Overview
Generally speaking, I think most of us would expect more TPE to equal more wins, it sounds logical right? So, is that the case?
Well, looking at the top achieving teams in this 6 season period, maybe not. The most wins a team managed in a single season over this stretch was 13, this was achieved on 4 different occasions: S44 CTC, S46 AZ, S47 SJS and S49 YKW. Looking at the TPE metrics, I can only conclude that 13 win teams are anomalous as the 4 teams tied at the top of the wins column rank the following for TPE:
S44 CTC - 17th
S47 SJS - 21st
S46 AZ - 22nd
S49 YKW - 36th
Furthermore, while generally, more TPE does equal more wins, this is far from guaranteed and there are other outliers in the dataset:
For example, both S45 OCO and S44 BER had over 20,000 total TPE, 2 of 9 teams in this 6 season span to have >20,000. However, while the other 7 teams in this bucket achieved a minimum of 10 wins, with most actually recording 11 or 12, both S45 OCO and S44 BER only achieved 8 wins.
If we look at the dataset with the 13 win teams removed as outliers, there were 24 teams that won double digit games (10, 11 or 12 wins).TPE-wise, these teams can be separated into the following groups:
- 7 had total TPE greater than 20000
- 12 had total TPE between 17500 and 20000
- 5 had total TPE between 15000 and 17500
In summary, of the 24 teams that achieved 10+ wins in a single season over this 6 season period, they all had in excess of 15000 Total TPE, and 19 of the 24 had greater than 17500 Total TPE.
Interestingly, of the 12 win teams, not a single one had a Total TPE under 17500. It is worth noting that while 4 teams achieved 12 wins with between 17500 and 20000 Total TPE, there were 20 total teams in this range, who won anywhere from 4 to 12 games.
The biggest underachievers in this period were S46 NOLA, riding a total TPE of 18,392 (25th highest in this period) to just 4 wins, a bottom 12 record.
S44 SJS also stand out as underachievers here, recording the lowest win total (3) for a team with over 15000 total TPE (15896). Only 5 of the 33 teams over this period with over 15000 Total TPE failed to record 6 or more wins.
On the other end of the spectrum, we had a number of teams over the duration of this dataset who we can deem overachievers. These are teams who TPE wise, do not stand out from the crowd, or are downright bad, but outperformed expectations. Overachievers include:
- S49 NYS - 3rd lowest total TPE over the last 6 years, managed to win 7 games
- S49 YKW - Virtually league average (16824 vs 16589) total TPE, 13 wins (This felt worthy of a separate call out given the especially average Total TPE number, despite all 13 win teams being, inherently, massive overachievers)
- S47 and S48 OSK - 11 wins per season, 17022 and 15933 total TPE (second one of these is actually below league average)
- S45 SAR - only team under 15000 total TPE to finish over 0.500 (won 9 games)
There are a couple of franchises I would like to give a shout out to at this point, albeit begrudgingly. While some teams have overachieved, and others have underachieved, I think one of the most impressive stats I came across in this dataset is the consistency shown by the Baltimore Hawks and the Arizona Outlaws over the last 6 seasons.
BAL and AZ are the only franchises over the last 6 seasons to have all 6 rosters in the top half for total TPE, with the lowest of the 12 coming in at 32nd of 84. This means that between them, BAL and AZ make up 28.6% of the top half of rosters over this period - that is genuinely impressive.Interestingly, despite their consistency, none of the BAL or AZ rosters feature in the top 10 for total TPE over this period, with the highest being S47 BAL coming in at 12th.
This consistency in TPE is matched by consistency in record, as over this 6 season stretch, these 12 teams all won a minimum of 9 games, averaging 10.6 wins a season. A fun little data point I did notice when looking at BAL and AZ rosters over the last 6 seasons is that 10 of the 12 rosters featured 0 TPE at TE - BAL had 2 teams with TPE at TE, 276 in S46 and 679 in S45.
Correlation
Once all the data was collected, I extracted what I needed from the master table in order to produce graphs that show the direct correlation (or lack thereof) of TPE against a number of metrics that contribute to, or define, team success.
Disclaimer: Correlation does not equal causation - please don’t shout at me Tmoney
Wins
In a regular season focused study, wins are inarguably the most important metric when defining success.
Over the last 6 seasons, teams in the ISFL have won anywhere between 0 and 13 games each season, with the most common win total being 9 wins (13 teams). 4 teams share the record for win total in this period, and only 1 team failed to record a single win.
Fun little quirk in the data, exactly 8 teams have won 10, 11 and 12 games in the past 6 seasons. Not sure why 8 is the magic number, but I wanted to point it out regardless.
The following section contains graphs that show how correlated a specific TPE metric is to win total. Included in the figures is an R-squared value, also called the coefficient of determination, and gives a value between 0 and 1.0 as a measure of the ‘goodness of fit’ of a data set. In English, this means that the closer the R-squared number is to 1.0, the more correlation there is between the TPE metric and number of wins.
As a general rule of thumb:
<0.3 - No effect
0.3 - 0.5 - Weak effect
0.5 - 0.7 - Moderate effect
>0.7 - Strong effect
Wins vs TOTAL TPE
As we can see from the graph, there is a positive correlation between TOTAL TPE and wins in the ISFL - I know, this is groundbreaking stuff.
An R-squared value of 0.577 indicates there is a moderate effect of TOTAL TPE on Wins, meaning that as TOTAL TPE increases, so does the recorded Win total.
![[Image: WJRT9dy.jpeg]](https://imgur.com/WJRT9dy.jpeg)
Given the 13 win teams were largely outliers, I was curious as to how this relationship changed if they were removed from the data set. The following graph shows this, TOTAL TPE vs Wins with 13 win teams excluded. Interestingly enough, the fit is better, with the R-squared value increasing to 0.594. While I appreciate this is not an overwhelmingly significant change, it does further support the notion that 13 win teams are largely outliers (note I said outliers, and not flukes, this is a scientific study).
![[Image: 3uSPcvq.jpeg]](https://imgur.com/3uSPcvq.jpeg)
Wins vs Number of players >1000 TPE
I don’t know if it’s just me or if this is a widely held view, but there’s something about the number 1000. I can’t explain it, maybe it’s simple monkey brain math of ‘MORE DIGITS BETTER’ (yes, I even said math for you lot, you trans-atlantic heathens).
Despite being such a highly coveted number, a 1000 TPE player isn’t all that rare. In S49 there were 66 players with over 1000 TPE as of Week 1 of the regular season, of 326 players total. Over 20% of all players in the ISFL in S49 were over 1000 TPE, I thought that was surprisingly high - especially when you take into consideration the fact that a number of other players who weren’t at this magical threshold will have surpassed it over the course of the season.
So, the question remains, does having more players over 1000 TPE equate to more wins? In a word, not really.
Of all the team TPE metrics I looked at in this study, this was actually the least correlated with record. With an R-squared value of just 0.357, there is at most a weak link between the number of players over 1000 TPE on roster and win total. This is further evidenced by looking at the data itself as, for example, 12 win teams have anywhere between 3 and 13 players over 1000 TPE, and teams with 4 wins or fewer have anywhere from 0 to 7. Only 1 team in the past 6 seasons failed to have a player with over 1000 TPE on the roster for Week 1 of the regular season, S44 NOLA.
![[Image: k9rG24h.jpeg]](https://imgur.com/k9rG24h.jpeg)
Wins vs Average TPE
Finally, I wanted to look at whether the average TPE of players on active rosters had an impact on win total. This was an attempt to remove a limitation of looking at TOTAL TPE as, for example, a team with 5 WRs of 600 TPE each may have different success to a team with 3 WRs of 1000 TPE, however the total TPE would be the same.
I debated whether to even bother collecting this data while pouring over depth charts and TPE tracker graphs, but I’m very glad I decided to include it.
Looking at the below graph, we can see that Avg TPE is the most positively correlated team TPE metric to win total, with an R-squared value of 0.585. This is the highest R-squared value we have seen when looking at the total data set.
![[Image: CTQdqmZ.jpeg]](https://imgur.com/CTQdqmZ.jpeg)
When we remove the 13 win teams from the data set as an outlier, as we did with TOTAL TPE, the correlation becomes even stronger, creeping over 0.6 to give us an R-squared value of 0.607. If we think about what this means in a practical sense, the findings make sense. It would be quite easy to 'pad' a Total TPE stat with sheer volume of players, but, given that the game of football is won on 1-to-1 matchups a lot of the time, these teams may struggle to get over the line as individual TPE numbers would be generally poor. These R-squared values show that, while having a lot of TPE is generally good, when building a roster it is as important (if not more given our findings earlier on >1000 TPE players) to make the bottom of your TPE tracker stronger if you want to compete.
![[Image: DjpAInr.jpeg]](https://imgur.com/DjpAInr.jpeg)
Extrapolating
In theory, if this data is accurate and actually tells us anything, we should be able to look forward to the coming season and extrapolate from this data set to predict how teams will do in S50. Following updates over the weekend, the TPE tracker is ripe with fresh, juicy numbers for me to pluck ahead of tonight's Week 1 matchups.
Disclaimer: Yes, this says ahead of Week 1, this was the initial plan but unfortunately things got delayed. You have my sincere apologies.
S50 TPE Totals: Total TPE, OFF TPE and DEF TPE
![[Image: 8WHshKn.jpeg]](https://imgur.com/8WHshKn.jpeg)
ASFC
Looking at the table above, the standout roster (once again) is NOLA, the only team above 20000 and clearing the rest of the league by over 4500 Total TPE. They also top both individual rankings, being the only roster with over 10000 OFF TPE and over 11000 DEF TPE. There are some ominous early signs of a NOLA three-peat on the cards, as I’m genuinely not sure who is going to be able to stop them looking at the TPE data. The best hope for the league at this point appears to be a prayer to the sim gods.
Predicted win total: 14
I know, I know, I said all 13+ win seasons were a fluke. But I just can’t see NOLA not dominating this season, and I think the win total reflects that.
HONs spell at the bottom appears to be over, coming into S50 with the 3rd highest Total TPE in the entire league and the number 2 in their conference. While HON are still catching up with the big boys on the defensive side of the ball, their offense has developed nicely over the last few seasons and come into S50 as the 2nd ranked unit. Based on the TPE alone, HON could see a return to the playoffs this season after a long hiatus.
Predicted win total: 11
SJS are definitely not at the TPE heights they were at a few seasons ago, but they still come in a 4th for Total TPE. A strong looking offense, especially RB room, will look to take some of the pressure off a regression-hit Patterson, and they will hope a strong ground game will make up for an average TPE total of the defensive side of the ball. I think the 3rd playoff spot in the ASFC is going to come right down to the wire, and I wouldn’t like to call it at this stage.
Predicted win total: 10
An alien environment for AZ this season, as they find themselves slap bag in the middle of average when it comes to TPE. A total of 16860 is good enough for 4th in the division, and 6th in the league. AZ will be hoping they can keep tabs on SJS in the rush for the third playoff spot, as they look to prove they are ‘best of the rest’. Regression has hit a number of big names on the AZ roster, and their defense looks rather depleted heading into S50, as AZ come into the season with a bottom 3 unit in the league.
Predicted win total: 10
I know, this is still a lot of wins given the TPE numbers, but the rest of this conference looks so bad that someone is going to have to win more games than they ‘should’.
AUS come into the season in a bit of a strange position, a little afloat from the other ‘middle of the pack’ teams, but with a good 1000+ TPE cushion before the next roster behind them. The Copperheads will look towards a strong defensive unit (top 4 in the league), to keeps games tight so their lacklustre offense doesn’t have too much work to do. I expect AUS to sneak some wins in games where their defense can stay in control, but if they find themselves in shoot outs I can see them struggling.
Predicted win total: 6
NYS over-performed massively last season and, from a TPE perspective at least, they did not get better. I can see this being a long season for the Silverbacks, as they look weak on both sides of the ball, coming in in the bottom 4 for both OFF TPE and DEF TPE. Sure, I can see them nicking the odd win here and there, but I don't think they’ll be favoured in many matchups.
Predicted win total: 3
OCO have been trending this way for a few seasons now, after the heights of S45-48, and it looks like this season they finally hit the ‘fuck it’ button. Having traded away a couple of developing players to contenders, OCOs roster looks rough this season, starting S50 with the lowest DEF TPE in the league by over 2000 TPE, and the 2nd lowest OFF TPE in the league as well. They are bottom for both in their conference, and are headed into week 1 with the 2nd lowest Total TPE we have seen in the last 7 seasons. I am hard pressed to find a game on their schedule I can favour them in.
Predicted win total: 0
NSFC
Osaka sit at the top of the pile in the NSFC, and I have the win total reflecting the same. As the only team in the conference with over 18000 Total TPE, and the only team with over 8000 OFF TPE, OSK will rarely be out of contention at the end of games. While they don’t top the charts for DEF TPE, over 9000 is still nothing to scoff at and should be more than enough to get them over the line more often than not.
Predicted win total: 12
BAL come into the season with an unusual roster, given their recent preferences. Over the last few seasons the Hawks have been at, or near, the top of the OFF TPE rankings, however in S50 it looks like their defense is going to have to do some of the heavy lifting. Regression has hit a few stars on the offensive side of the ball and they now find themselves in the middle of the pack, 6th in their conference and 9th in the league, for OFF TPE. They do, however, come in at number 3 league wide for DEF TPE, and should still be favoured in most matchups.
Predicted win total: 11
The start of the dog fight for the 3rd playoff spot in the NSFC is SAR. The likely favourites to take the final spot, given they are the only remaining team in the conference with over 17000 Total TPE, SAR will be hoping that once again they can overachieve on offense. SAR are no strangers to lacklustre OFF TPE numbers and a strong defense, but in the past they have consistently put up more PF than expected. Will they do the same in S50, or will they regress to the mean and fall short? I have them just holding on, but the middle 4 teams in this conference could honestly go in any direction.
Predicted win total: 9
Separating the next 3 teams was difficult, really difficult, and as a result, I didn’t bother:
YKW were an outlier last time out (I’m sure my feelings on 13 win teams have become apparent at this point) and I cannot see them repeating that task. From a TPE perspective, YKW have got worse since last season, but they have made a couple of moves this offseason that could pay dividends down the stretch. Although YKW are top of these 3 in terms of total TPE, they are bottom of the 3 in both OFF TPE and DEF TPE, with the difference being made up on special teams. As a result of this, if a team is to fall off from the pack I can see it being YKW.
CTC come into the season as THE middle of the pack team. They rank 9th in total TPE, 7th in OFF TPE and 6th in DEF TPE. Cape town look stronger on offense than defense this season, and with a QB reaching his peak they will be hoping that will be the difference between scraping a playoff berth and disappointment. CTC come into the season as the 3rd ranked offense in the conference TPE wise.
Last but certainly not least of these 3 teams, BER may come into the season with the 2nd lowest total TPE in the conference, but they are only just over 1000 TPE off BAL in 2nd. This conference could genuinely finish in any arrangement of these teams and it wouldn’t be all that surprising, and a young BER team will be hoping they can take advantage of the chaos to sneak into the playoffs. While not necessarily likely, I can see BER riding the 2nd highest OFF TPE to a couple of extra upset wins and sneaking into the 3rd spot (I am not biased at all I promise).
Predicted win total: 8
I’m not sure this will come as a shock to anyone, as although there is only 2000 TPE between the other 6 teams in the conference, COL find themselves adrift at the bottom by nearly 4000 TPE. While the standings leave much to be desired for COL, the numbers themselves are promising, with the highest Total TPE for a Yeti roster in the past 4 seasons. It is on offense where I think COL will struggle, with the lowest OFF TPE in the entire league. I can see them sneaking a couple of wins, especially against a few of the ASFC teams, but anymore than a couple of victories would be a massive over-achievement.
Predicted win total: 2
Hope you enjoyed the first instalment of this deep dive into what is proving to be quite an interesting data set so far. I had initially intended to do this as one media piece but it’s ended up being too much for the one piece. So, stay tuned for the next part where I will be diving into TPE allocated to offense vs defense, different position groups, and the correlation of all of these numbers with win totals and PF/PA.
If anyone made it this far thanks for reading, I really appreciate it! I wanted to get the first instalment of this done in time for 2x media but just fell short: sometimes you win, sometimes you lose.
Any sleights, perceived or otherwise, against specific teams or player groups are not intended to offend and are merely either what the data shows (data doesn’t lie) or ill-conceived attempts at humour (I blame the culture barrier).
Introduction
So, I guess the first question I need to answer is: why? Why have I decided to scour the rosters of the past 6 seasons for TPE totals? The truth of the matter is, I was starting a different media project that required data on teams TPE allocation to different position groups. After being bounced around various senior and long-standing league members, I unfortunately came to the conclusion that the data I was after did not exist. This left me with 2 choices; either I give up on the media project as a whole, or I sack it up and collect/process the data myself. At this point it was suggested to me that I could automate this process, but I unfortunately have to admit that my computer skills shame my generation and are more in line with those expected of a 60 year old, and are certainly not able to complete such a task.
Fortunately for me, I am used to brainlessly and stubbornly brute forcing my way through large amounts of data for league media, having produced a number of statistically centred media pieces early in my league career.
I started by using the ISFL index to record the rosters for all ISFL teams over the last 6 seasons: S44, S45, S46, S47, S48 and S49. I then used the schedules from each season to determine the latest date a player could be updated before week 1 of each season. I'm aware that players can earn TPE over the season so this may not be the most accurate representation of TPE, but I felt it made sense as the baseline for a rosters TPE as otherwise I felt like it would become too complicated to keep track of. The dates for prior to week 1 updates were as follows:
S44 - Oct 30th
S45 - Nov 25th
S46 - Jan 27th
S47 - Mar 23rd
S48 - May 18th
S49 - Jul 13th
Once I had a comprehensive list of players for each season, I was able to cross reference this with the above dates in the TPE tracker, in order to record each player's TPE on week 1 of each listed season. Once I had completed this, I had a complete roster with TPE for each team over the last 6 seasons, accurate as of week 1 of each relevant season.
Yes, this was time consuming, it has cost me my social life, distracted me at work and put a strain on a number of relationships over the last few weeks. When I close my eyes I see depth charts, I have TPE numbers and week 1 dates keeping me awake at night, I have player career paths embedded into my brain that went IA 3 seasons ago - I may need help.
Ultimately, this then allowed me to get an overview of which teams had how much TPE allocated to each position group over the last 6 seasons, and how this aligned with their success in the regular season, in the form of win totals. I decided to only use the regular season for ease of data extraction, and also because we all know that the playoffs are a crap shoot and the sim gods toy with us - I am not bitter at all, I promise.
For simplicity's sake, I have labelled all teams that were a part of the Chicago/Osaka franchise as OSK, regardless of whether they existed before or after the re-brand.
General Impressions
The first thing I did was produce ISFL TPE heat maps for each team in each of the last 6 seasons. I will attach these below, and am aiming to complete the same in the coming seasons so this isn't as big a task should anyone want to look into this kind of data again.
Once I had the TPE heat maps, and a master table, completed I was able to start analyzing the data. I started by combing through the raw data and pulling out any data points that stood out to me, before I then went on to process the data more formally. I am going to start by discussing the data points that stood out to me on first pass.
The TPE metrics I extracted from the data set where as follows:
Total TPE - sum of TPE of all players on a roster in a given season
OFF TPE - sum of TPE of all players on offense in a given season
DEF TPE - sum of TPE of all players on defense in a given season
ST TPE - sum of TPE of all players on special teams in a given season
QB TPE - sum of TPE of all players assigned QB in a given season
RB TPE - sum of TPE of all players assigned RB in a given season
WR TPE - sum of TPE of all players assigned WR in a given season
TE TPE - sum of TPE of all players assigned to TE in a given season
OL TPE - sum of TPE of all players assigned to OL in a given season (this includes T, G and C)
DL TPE - sum of TPE of all players assigned to DL in a given season (this includes both DE and DT)
LB TPE - sum of TPE of all players assigned to LB in a given season
DB TPE - sum of TPE of all players assigned to DB in a given season (this includes both CB and S)
>1000 TPE - number of players on a roster over 1000 TPE, regardless of position
Avg TPE - TOTAL TPE divided by total number of players on the roster in a given season
I only included TPE allocated to user controlled players, not bots. This included both active players and IA players, regardless of whether drafted or signed via FA.
The raw data can be found at the below links if anyone is interested:
Rosters:
https://docs.google.com/spreadsheets/d/1...drive_link
Master Table - S44-S49:
https://docs.google.com/spreadsheets/d/1...drive_link
Averages
The first piece of data I wanted to look at was an average, as I would therefore then have something to compare the other numbers to. The means are as follows:
- Total 16589
- OFF 6953
- DEF 9083
- QB 1064, RB 1687, WR 2344, TE 395, OL 1467
- DL 2163, LB 2912, DB 4009
- ST 548
- >1000 players 5, average TPE per ISFL player 716
The modal ranges are as follows:
- Total 15000 - 17500
- OFF 6000 - 8000
- DEF 8000 - 10000
- QB 1000 - 1200, RB 1500 - 2000, WR 2500 - 3000, TE 0, OL 1000 - 2000
- DL 2000 - 2500, LB 2500 - 3000, DB 3000 - 5000
- ST 250 - 500
- >1000 players, average TPE per ISFL player
Total TPE
I'm going to start by talking about the top of the dataset, and a few teams that stood out to me TPE-wise across this period:
NOLA:
One of the repeated names at the top of the rankings or Total TPE are the New Orleans Second Line, also referred to as NOLA. When looking at the last 6 seasons, a few periods of dominance stand out, and S47-49 NOLA is perhaps the best of the bunch. When including data from the past 6 seasons, NOLA take 3 of the top 4 spots for total TPE with their rosters in the 3 most recent seasons. This spell at the top has lead to impressive win totals, as well as back to back championships in the last 2 seasons (I know I said playoffs don't matter, I'm a hypocrite, get over it).
NOLA have been a bit of a Cinderella story over the time period that falls under the scope of this study, reaching the aforementioned spell near the top of the rankings at the end of the assessed period, but starting in S44 with a bottom 10 total TPE for this entire 6 season stretch. The NOLA roster in S44 is also the only roster in this dataset to fail to record a single player with 1000 or more TPE on week 1 of the season.
When going back through the index to construct the rosters, one thing in particular stood out to me about NOLA in the 6 season stretch - they have consistently had low roster turnover season to season, losing and replacing a maximum of 1 or 2 players a season. This has allowed the roster to earn and grow together, with the same nucleus of the team present in S44, when they had 0 players over 1000 TPE, right through to S48/49 when they had 11 players over 1000 TPE. While this approach to roster building has clearly worked, it does leave you wondering where they'll be when this nucleus enters regression. Could we see NOLA back at the bottom of the pile in the next few seasons?
Other notable periods of dominance over this time period include S45-47 OCO and S44-45 HON. These 3 teams (S47-49 NOLA, S45-47 OCO and S44-45 HON) make up the top 8 total TPE rosters over the last 6 seasons. These 8 teams won a total of 88 games, averaging 11 wins per season over this stretch. For those looking for an early answer to ‘does TPE matter’, the early evidence certainly looks promising.
A few things I noticed in common for these 8 teams:
- All have over 1000 TPE allocated at QB
- All have over 2500 TPE allocated to WRs
- 6/8 teams have more than 3000 TPE allocated to OL
- All 5 non-OCO teams all have ~3000+ TPE at DL
- General trend is less TPE at RB and DB - 4/8 teams have under league average TPE at RB
NOLA stand out amongst top 8 for having a lot of TPE at RB - only teams to have over 2000 TPE at RB are both NOLA (S48 and S49)
OCO:
In the early portion of this 6 season stretch, the Orange County Otters (also called OCO) were frequently at the top of the rankings, both in terms of TPE and record. In fact, S45, S46 and S47 OCO rosters are 3 of the 9 rosters to start a season with over 20000 Total TPE in this period. However, despite consistent appearances at the top of the total TPE lists across this period, OCO didn’t build their rosters the same as everyone else.
The first thing I noticed with OCO, especially with their successful teams, was a relative lack of TPE allocated to DL. Despite their success in these seasons, OCO were twice in the bottom 11 of DL TPE over this period - 1461 TPE in S45 and 1392 in S46. These numbers are both well below league average at DL, and make their successes that much more impressive. Their successes on defense in spite of this may be explained by the measures taken to mitigate this potential weakness. S46 OCO had the highest TPE at LB over this 6 season period at 4953, and OCO S45 had the highest TPE at DB over the last 6 seasons at 6275. Whether this was an intentional approach or not, making the rest of the roster on defense some of the strongest groups we have seen in the last 6 years certainly won’t have hurt their attempts to mitigate their weakness on the DL. Interestingly, the only other team to have more than 6000 TPE invested at DB in this period is S46 OCO.
It wasn’t just on defense where OCO went against the grain. S46 OCO were second in Total TPE over this period but are the only team in top 10 for total TPE in the past 6 seasons to be outside the top 50 for TPE at RB position. OCO noticeably invested TPE on the OL, with both S46 and S45 OCO recording over 3000 TPE to start the season (2 of 8 total teams) and S47 narrowly outside (2912), meaning OCO take one third of the top 9 spots for OL TPE in this period. Not many teams prioritised OL TPE, so this data point stood out to me.
HON:
The Honolulu Hahalua, also referred to as HON, were almost the reverse NOLA over this time period, starting off very strong in S44 and S45 and then undergoing a dramatic rebuild *cough* tank *cough* in S46. To illustrate just how dramatic this change was, S44 and S45 HON appear in the top 8 for total TPE. HON then fail to appear in the top 50 for Total TPE over the following FOUR seasons, achieving a high of 55 with S49. The other 3 HON rosters (S46, S47 and S48) are all in bottom 15 for total TPE over the assessed period.
An interesting data point from the Hahalua, S44 and S45 HON are only teams in top 11 total TPE to have 0 TPE allocated to the TE position (more on this later).
Moving on, unfortunately for some of you reading this, it is time to talk about the bottom of the pile:
COL:
The Colorado Yeti, or COL, have fallen on hard times recently, and it is probably not a surprise to anyone that has paid attention to rosters, win totals, or the draft over the last couple of seasons to see them at the bottom of the table when it comes to looking at TPE and wins.
COL have the 2 lowest, and 3 of the bottom 5, total TPE rosters in the past 6 seasons (S47, S48 and S49). S47 COL are the only team in the last 6 seasons to record less than 11000 Total TPE on a roster.
Some of you may have already noticed this but yes, this means that 13 teams defenses, and S48 NOLAs offense, had more TPE as a unit than S47 COLs whole roster (sorry COL, but this data point just cannot be ignored).
One thing that stood out to me when looking at COLs rosters was the notably low TPE at WR. S47 and S48 COL are the only teams in the last 6 seasons to have less than 1000 TPE total at WR.
The low TPE totals are reflected in the records of these teams, most notably as S47, S48 and S49 COL are the only teams to have won less than 2 games in a season over this period.
I do want to give a shoutout to COL GM Aeonsjenni for having the only 1000 TPE player on COL at the start of each of the past 3 seasons.
NYS:
The other team to join COL at the bottom of the rankings are NYS, who claim the other 2 spots in the bottom 5 for Total TPE over the past 6 seasons. Despite having posted 11 wins over these 2 seasons, S48 and S49 NYS are the 3rd and 4th lowest total TPE rosters of the past 6 seasons. This makes their win total that much more impressive, as NYS have no single green position group on the roster through S48 and S49 except for K/P.
One other thing I noticed when looking at the rosters at the bottom of the table for Total TPE was that the bottom 5 teams have notably very little TPE allocated to OL. The league average for TPE across the OL group as a whole is 1467, but the totals for the bottom 5 teams are 0, 261, 0, 171 and 228. I know the league generally struggles to find active earners on the OL, and it may just be that during a rebuild teams prefer to use their active players at positions they deem more valuable, but I thought it was an interesting data point regardless.
Total TPE vs wins - Overview
Generally speaking, I think most of us would expect more TPE to equal more wins, it sounds logical right? So, is that the case?
Well, looking at the top achieving teams in this 6 season period, maybe not. The most wins a team managed in a single season over this stretch was 13, this was achieved on 4 different occasions: S44 CTC, S46 AZ, S47 SJS and S49 YKW. Looking at the TPE metrics, I can only conclude that 13 win teams are anomalous as the 4 teams tied at the top of the wins column rank the following for TPE:
S44 CTC - 17th
S47 SJS - 21st
S46 AZ - 22nd
S49 YKW - 36th
Furthermore, while generally, more TPE does equal more wins, this is far from guaranteed and there are other outliers in the dataset:
For example, both S45 OCO and S44 BER had over 20,000 total TPE, 2 of 9 teams in this 6 season span to have >20,000. However, while the other 7 teams in this bucket achieved a minimum of 10 wins, with most actually recording 11 or 12, both S45 OCO and S44 BER only achieved 8 wins.
If we look at the dataset with the 13 win teams removed as outliers, there were 24 teams that won double digit games (10, 11 or 12 wins).TPE-wise, these teams can be separated into the following groups:
- 7 had total TPE greater than 20000
- 12 had total TPE between 17500 and 20000
- 5 had total TPE between 15000 and 17500
In summary, of the 24 teams that achieved 10+ wins in a single season over this 6 season period, they all had in excess of 15000 Total TPE, and 19 of the 24 had greater than 17500 Total TPE.
Interestingly, of the 12 win teams, not a single one had a Total TPE under 17500. It is worth noting that while 4 teams achieved 12 wins with between 17500 and 20000 Total TPE, there were 20 total teams in this range, who won anywhere from 4 to 12 games.
The biggest underachievers in this period were S46 NOLA, riding a total TPE of 18,392 (25th highest in this period) to just 4 wins, a bottom 12 record.
S44 SJS also stand out as underachievers here, recording the lowest win total (3) for a team with over 15000 total TPE (15896). Only 5 of the 33 teams over this period with over 15000 Total TPE failed to record 6 or more wins.
On the other end of the spectrum, we had a number of teams over the duration of this dataset who we can deem overachievers. These are teams who TPE wise, do not stand out from the crowd, or are downright bad, but outperformed expectations. Overachievers include:
- S49 NYS - 3rd lowest total TPE over the last 6 years, managed to win 7 games
- S49 YKW - Virtually league average (16824 vs 16589) total TPE, 13 wins (This felt worthy of a separate call out given the especially average Total TPE number, despite all 13 win teams being, inherently, massive overachievers)
- S47 and S48 OSK - 11 wins per season, 17022 and 15933 total TPE (second one of these is actually below league average)
- S45 SAR - only team under 15000 total TPE to finish over 0.500 (won 9 games)
There are a couple of franchises I would like to give a shout out to at this point, albeit begrudgingly. While some teams have overachieved, and others have underachieved, I think one of the most impressive stats I came across in this dataset is the consistency shown by the Baltimore Hawks and the Arizona Outlaws over the last 6 seasons.
BAL and AZ are the only franchises over the last 6 seasons to have all 6 rosters in the top half for total TPE, with the lowest of the 12 coming in at 32nd of 84. This means that between them, BAL and AZ make up 28.6% of the top half of rosters over this period - that is genuinely impressive.Interestingly, despite their consistency, none of the BAL or AZ rosters feature in the top 10 for total TPE over this period, with the highest being S47 BAL coming in at 12th.
This consistency in TPE is matched by consistency in record, as over this 6 season stretch, these 12 teams all won a minimum of 9 games, averaging 10.6 wins a season. A fun little data point I did notice when looking at BAL and AZ rosters over the last 6 seasons is that 10 of the 12 rosters featured 0 TPE at TE - BAL had 2 teams with TPE at TE, 276 in S46 and 679 in S45.
Correlation
Once all the data was collected, I extracted what I needed from the master table in order to produce graphs that show the direct correlation (or lack thereof) of TPE against a number of metrics that contribute to, or define, team success.
Disclaimer: Correlation does not equal causation - please don’t shout at me Tmoney
Wins
In a regular season focused study, wins are inarguably the most important metric when defining success.
Over the last 6 seasons, teams in the ISFL have won anywhere between 0 and 13 games each season, with the most common win total being 9 wins (13 teams). 4 teams share the record for win total in this period, and only 1 team failed to record a single win.
Fun little quirk in the data, exactly 8 teams have won 10, 11 and 12 games in the past 6 seasons. Not sure why 8 is the magic number, but I wanted to point it out regardless.
The following section contains graphs that show how correlated a specific TPE metric is to win total. Included in the figures is an R-squared value, also called the coefficient of determination, and gives a value between 0 and 1.0 as a measure of the ‘goodness of fit’ of a data set. In English, this means that the closer the R-squared number is to 1.0, the more correlation there is between the TPE metric and number of wins.
As a general rule of thumb:
<0.3 - No effect
0.3 - 0.5 - Weak effect
0.5 - 0.7 - Moderate effect
>0.7 - Strong effect
Wins vs TOTAL TPE
As we can see from the graph, there is a positive correlation between TOTAL TPE and wins in the ISFL - I know, this is groundbreaking stuff.
An R-squared value of 0.577 indicates there is a moderate effect of TOTAL TPE on Wins, meaning that as TOTAL TPE increases, so does the recorded Win total.
Given the 13 win teams were largely outliers, I was curious as to how this relationship changed if they were removed from the data set. The following graph shows this, TOTAL TPE vs Wins with 13 win teams excluded. Interestingly enough, the fit is better, with the R-squared value increasing to 0.594. While I appreciate this is not an overwhelmingly significant change, it does further support the notion that 13 win teams are largely outliers (note I said outliers, and not flukes, this is a scientific study).
Wins vs Number of players >1000 TPE
I don’t know if it’s just me or if this is a widely held view, but there’s something about the number 1000. I can’t explain it, maybe it’s simple monkey brain math of ‘MORE DIGITS BETTER’ (yes, I even said math for you lot, you trans-atlantic heathens).
Despite being such a highly coveted number, a 1000 TPE player isn’t all that rare. In S49 there were 66 players with over 1000 TPE as of Week 1 of the regular season, of 326 players total. Over 20% of all players in the ISFL in S49 were over 1000 TPE, I thought that was surprisingly high - especially when you take into consideration the fact that a number of other players who weren’t at this magical threshold will have surpassed it over the course of the season.
So, the question remains, does having more players over 1000 TPE equate to more wins? In a word, not really.
Of all the team TPE metrics I looked at in this study, this was actually the least correlated with record. With an R-squared value of just 0.357, there is at most a weak link between the number of players over 1000 TPE on roster and win total. This is further evidenced by looking at the data itself as, for example, 12 win teams have anywhere between 3 and 13 players over 1000 TPE, and teams with 4 wins or fewer have anywhere from 0 to 7. Only 1 team in the past 6 seasons failed to have a player with over 1000 TPE on the roster for Week 1 of the regular season, S44 NOLA.
Wins vs Average TPE
Finally, I wanted to look at whether the average TPE of players on active rosters had an impact on win total. This was an attempt to remove a limitation of looking at TOTAL TPE as, for example, a team with 5 WRs of 600 TPE each may have different success to a team with 3 WRs of 1000 TPE, however the total TPE would be the same.
I debated whether to even bother collecting this data while pouring over depth charts and TPE tracker graphs, but I’m very glad I decided to include it.
Looking at the below graph, we can see that Avg TPE is the most positively correlated team TPE metric to win total, with an R-squared value of 0.585. This is the highest R-squared value we have seen when looking at the total data set.
When we remove the 13 win teams from the data set as an outlier, as we did with TOTAL TPE, the correlation becomes even stronger, creeping over 0.6 to give us an R-squared value of 0.607. If we think about what this means in a practical sense, the findings make sense. It would be quite easy to 'pad' a Total TPE stat with sheer volume of players, but, given that the game of football is won on 1-to-1 matchups a lot of the time, these teams may struggle to get over the line as individual TPE numbers would be generally poor. These R-squared values show that, while having a lot of TPE is generally good, when building a roster it is as important (if not more given our findings earlier on >1000 TPE players) to make the bottom of your TPE tracker stronger if you want to compete.
Extrapolating
In theory, if this data is accurate and actually tells us anything, we should be able to look forward to the coming season and extrapolate from this data set to predict how teams will do in S50. Following updates over the weekend, the TPE tracker is ripe with fresh, juicy numbers for me to pluck ahead of tonight's Week 1 matchups.
Disclaimer: Yes, this says ahead of Week 1, this was the initial plan but unfortunately things got delayed. You have my sincere apologies.
S50 TPE Totals: Total TPE, OFF TPE and DEF TPE
ASFC
Looking at the table above, the standout roster (once again) is NOLA, the only team above 20000 and clearing the rest of the league by over 4500 Total TPE. They also top both individual rankings, being the only roster with over 10000 OFF TPE and over 11000 DEF TPE. There are some ominous early signs of a NOLA three-peat on the cards, as I’m genuinely not sure who is going to be able to stop them looking at the TPE data. The best hope for the league at this point appears to be a prayer to the sim gods.
Predicted win total: 14
I know, I know, I said all 13+ win seasons were a fluke. But I just can’t see NOLA not dominating this season, and I think the win total reflects that.
HONs spell at the bottom appears to be over, coming into S50 with the 3rd highest Total TPE in the entire league and the number 2 in their conference. While HON are still catching up with the big boys on the defensive side of the ball, their offense has developed nicely over the last few seasons and come into S50 as the 2nd ranked unit. Based on the TPE alone, HON could see a return to the playoffs this season after a long hiatus.
Predicted win total: 11
SJS are definitely not at the TPE heights they were at a few seasons ago, but they still come in a 4th for Total TPE. A strong looking offense, especially RB room, will look to take some of the pressure off a regression-hit Patterson, and they will hope a strong ground game will make up for an average TPE total of the defensive side of the ball. I think the 3rd playoff spot in the ASFC is going to come right down to the wire, and I wouldn’t like to call it at this stage.
Predicted win total: 10
An alien environment for AZ this season, as they find themselves slap bag in the middle of average when it comes to TPE. A total of 16860 is good enough for 4th in the division, and 6th in the league. AZ will be hoping they can keep tabs on SJS in the rush for the third playoff spot, as they look to prove they are ‘best of the rest’. Regression has hit a number of big names on the AZ roster, and their defense looks rather depleted heading into S50, as AZ come into the season with a bottom 3 unit in the league.
Predicted win total: 10
I know, this is still a lot of wins given the TPE numbers, but the rest of this conference looks so bad that someone is going to have to win more games than they ‘should’.
AUS come into the season in a bit of a strange position, a little afloat from the other ‘middle of the pack’ teams, but with a good 1000+ TPE cushion before the next roster behind them. The Copperheads will look towards a strong defensive unit (top 4 in the league), to keeps games tight so their lacklustre offense doesn’t have too much work to do. I expect AUS to sneak some wins in games where their defense can stay in control, but if they find themselves in shoot outs I can see them struggling.
Predicted win total: 6
NYS over-performed massively last season and, from a TPE perspective at least, they did not get better. I can see this being a long season for the Silverbacks, as they look weak on both sides of the ball, coming in in the bottom 4 for both OFF TPE and DEF TPE. Sure, I can see them nicking the odd win here and there, but I don't think they’ll be favoured in many matchups.
Predicted win total: 3
OCO have been trending this way for a few seasons now, after the heights of S45-48, and it looks like this season they finally hit the ‘fuck it’ button. Having traded away a couple of developing players to contenders, OCOs roster looks rough this season, starting S50 with the lowest DEF TPE in the league by over 2000 TPE, and the 2nd lowest OFF TPE in the league as well. They are bottom for both in their conference, and are headed into week 1 with the 2nd lowest Total TPE we have seen in the last 7 seasons. I am hard pressed to find a game on their schedule I can favour them in.
Predicted win total: 0
NSFC
Osaka sit at the top of the pile in the NSFC, and I have the win total reflecting the same. As the only team in the conference with over 18000 Total TPE, and the only team with over 8000 OFF TPE, OSK will rarely be out of contention at the end of games. While they don’t top the charts for DEF TPE, over 9000 is still nothing to scoff at and should be more than enough to get them over the line more often than not.
Predicted win total: 12
BAL come into the season with an unusual roster, given their recent preferences. Over the last few seasons the Hawks have been at, or near, the top of the OFF TPE rankings, however in S50 it looks like their defense is going to have to do some of the heavy lifting. Regression has hit a few stars on the offensive side of the ball and they now find themselves in the middle of the pack, 6th in their conference and 9th in the league, for OFF TPE. They do, however, come in at number 3 league wide for DEF TPE, and should still be favoured in most matchups.
Predicted win total: 11
The start of the dog fight for the 3rd playoff spot in the NSFC is SAR. The likely favourites to take the final spot, given they are the only remaining team in the conference with over 17000 Total TPE, SAR will be hoping that once again they can overachieve on offense. SAR are no strangers to lacklustre OFF TPE numbers and a strong defense, but in the past they have consistently put up more PF than expected. Will they do the same in S50, or will they regress to the mean and fall short? I have them just holding on, but the middle 4 teams in this conference could honestly go in any direction.
Predicted win total: 9
Separating the next 3 teams was difficult, really difficult, and as a result, I didn’t bother:
YKW were an outlier last time out (I’m sure my feelings on 13 win teams have become apparent at this point) and I cannot see them repeating that task. From a TPE perspective, YKW have got worse since last season, but they have made a couple of moves this offseason that could pay dividends down the stretch. Although YKW are top of these 3 in terms of total TPE, they are bottom of the 3 in both OFF TPE and DEF TPE, with the difference being made up on special teams. As a result of this, if a team is to fall off from the pack I can see it being YKW.
CTC come into the season as THE middle of the pack team. They rank 9th in total TPE, 7th in OFF TPE and 6th in DEF TPE. Cape town look stronger on offense than defense this season, and with a QB reaching his peak they will be hoping that will be the difference between scraping a playoff berth and disappointment. CTC come into the season as the 3rd ranked offense in the conference TPE wise.
Last but certainly not least of these 3 teams, BER may come into the season with the 2nd lowest total TPE in the conference, but they are only just over 1000 TPE off BAL in 2nd. This conference could genuinely finish in any arrangement of these teams and it wouldn’t be all that surprising, and a young BER team will be hoping they can take advantage of the chaos to sneak into the playoffs. While not necessarily likely, I can see BER riding the 2nd highest OFF TPE to a couple of extra upset wins and sneaking into the 3rd spot (I am not biased at all I promise).
Predicted win total: 8
I’m not sure this will come as a shock to anyone, as although there is only 2000 TPE between the other 6 teams in the conference, COL find themselves adrift at the bottom by nearly 4000 TPE. While the standings leave much to be desired for COL, the numbers themselves are promising, with the highest Total TPE for a Yeti roster in the past 4 seasons. It is on offense where I think COL will struggle, with the lowest OFF TPE in the entire league. I can see them sneaking a couple of wins, especially against a few of the ASFC teams, but anymore than a couple of victories would be a massive over-achievement.
Predicted win total: 2
Hope you enjoyed the first instalment of this deep dive into what is proving to be quite an interesting data set so far. I had initially intended to do this as one media piece but it’s ended up being too much for the one piece. So, stay tuned for the next part where I will be diving into TPE allocated to offense vs defense, different position groups, and the correlation of all of these numbers with win totals and PF/PA.
If anyone made it this far thanks for reading, I really appreciate it! I wanted to get the first instalment of this done in time for 2x media but just fell short: sometimes you win, sometimes you lose.
Player Page | Wiki Entry
**Norfolk Seawolves GM**
2x Gemini Media Award Nominee
2x Cetus Media Award Nominee
![[Image: wCtE5t6.png]](https://imgur.com/wCtE5t6.png)
![[Image: AKelSrR.jpeg]](https://imgur.com/AKelSrR.jpeg)
https://wiki.sim-football.com/view/Mac_Mannheim
![[Image: tzcPBbQ.png]](https://i.imgur.com/tzcPBbQ.png)
![[Image: kai71.gif]](https://sig.grumpybumpers.com/host/kai71.gif)
![[Image: Mike_Gesicki_Retro.jpg?ex=66672085&is=66...height=670]](https://media.discordapp.net/attachments/758471909196562442/1245947218913919048/Mike_Gesicki_Retro.jpg?ex=66672085&is=6665cf05&hm=9887a80e5f603d7a8c10bfc6a4df7cf474317ce78789d9b8b920caf2c03dccba&=&format=webp&width=1005&height=670)
![[Image: CROney3.gif]](https://sig.grumpybumpers.com/host/CROney3.gif)
![[Image: 002p.png]](https://i.postimg.cc/nzK75H2c/002p.png)
