05-21-2024, 02:10 PM
(This post was last modified: 05-22-2024, 10:17 PM by wetwilleh. Edited 1 time in total.)
Hello everyone, I’ve got some fun historical ISFL team data to share, all courtesy of me being bored while waiting for the Timberwolves/Nuggets game! We’ll be looking into a dive of Glicko2 rating (ie. how good a specific team is) of every ISFL team, and Ultimus Overperformer/Underperformers!
First, a little technical bit on how I got the data. I recently started scraping the ISFL Index for Win/Loss data. Initially I was just doing it to see what scraping was like, I never really do it and so it was a fun project and I thought I could make an Elo thing but historical. I found that it was pretty easy to do, and that led me to having a data set (a big JSON file) of every single game that has been played in ISFL history (the ones found here: https://index.sim-football.com/ISFLS47/GameResults.html) and I thought that that was neat.
I was kinda annoyed with the shape of the data so I ended up just putting it all in a database and made a “dynamic” searchable API for anyone who might want the data for media of their own. It’s a fun little thing and I’ll add details at the end for anyone who wants to play with it. Let me know if there’s any other data scraping you would be interested in, I’m just going to build a big data thing so people can write media easier or satisfy their curiosity.
Anyway, with all this data, I realized I could do a historical Elo calculation of all teams. Elo is a rating system that tells you how good a team is based on how you win/lose to other teams. However, there’s a “better” system called Glicko (and Glicko2 as an improvement on that). Took me an hour to write the code for the info we have today.
What a Glicko rating is is a rating based on a set of matches, ie. 1 season. A team’s rating goes up or down based on who they beat within a season, but Glicko also has a built in mechanism for uncertainty on your rating if your team is winning or losing in an erratic manner (ie. Yellowknife these past seasons), or if your team is new to the dataset (the first 3-4 seasons of a team are kinda inaccurate because of model uncertainty).
TL;DR - Rating high = team good. First three seasons of a team's history are less accurate.
I used this data I had gotten programmatically to map the ratings of teams, and threw it all into a spreadsheet where we can make some cool charts and tables!
(https://docs.google.com/spreadsheets/d/1...sp=sharing)
Everyone loves charts! This is the history of the ISFL by rating. The average rating of the league is 1470.68 through all the seasons, and the median is about 1467. Notably, most teams in the league live around 1520-1410ish, and teams that go above or below that are pretty exceptional in one way or another. You’ll also see big swings the younger a team is, as the model tries to narrow down their certainty of the team’s performances.
Some key numbers:
So, if you’re a team competing for a championship, you want a rating of about +50 on the median, and conversely a bad team is about -50 on the median.
That chart above has way too much going on, so I’ve just copy and pasted the chart but only showing the ASFC teams. We’ll just go through each conference separately so we can actually see what’s happening.
The ASFC is characterized, generally, by a set of middling teams and then Orange County and Arizona standing way above them. This is going to be more apparent the further we go in this piece.
Arizona starts off the league strong with a three-peat under some “mysterious” circumstances (not mysterious, it was some “mild cheating”). After the multis get busted, Arizona becomes about an average ASFC team (which is a bad NSFC team) and starts improving around the Current Sim Era where they’ve reached pretty dominant success.
That Arizona dominance is really nothing compared to Orange County. The Otters are pretty unparalleled in the ISFL. I think the chart scale really does a disservice to how disgustingly good the Otters were for the first half of the league. Remember, the median league winner is a 1543 rating, the Otters went from S4-26 at 1560 rating, and were comfortably at 1600 for a long while. Of course, starting S27, they’ve slumped and have not had positive seasons since.
San Jose is our third Original 6 team, we’ve got some spikes at the fore-front of their history but they have only peaked at 1500+ once (S6) before being a 1440ish team, which is below the league median.
New Orleans (or the Las Vegas Legion) are our next team to join the ASFC. The Legion started out ROUGH. However, in S7, their fortunes started to change and they rocketed up the standings before turning into a league-average team and then starting to drop with the Current Sim Era and then of course the 40s started another huge down-tick. This past season however, sees a big rise, and we’re going to see them shoot up again as their low rating combined with wins will rocket them up.
Austin similarly started the league in a very rough space. No problem for this expansion team though, as they quickly built themselves as a competitor in the early 30s and managed to peak in the 40s as a championship team.
Honolulu joined the league in S22 and had a fantastic run in the mid 30s, clearly separating themselves from the rest of the ASFC before a tank, another rise, and then a sharp drop in the current seasons.
New York follows a trajectory that is pretty familiar by now, a rough start for an expansion team before quickly becoming a top squad for a few seasons, a partial rebuild trough, and then competition. This really is the ASFC, way more than the NSFC that we’ll see soon.
The NSFC is markedly different from the ASFC. First off, we don’t have that consistent 1600 rating squad that we saw, and also, the median is just higher. The NSFC teams, as a whole, are more consistent than the ASFC. The ASFC is marked by teams being really good or bad, and the NSFC has more parity within it.
Starting with Colorado, we see a team that struggled mightily at the start, coming to a nadir in S7, before a hard-fought battle before a strong S25-35 before dropping back down. Colorado is a pretty unique case, you don’t often see 10 season stretches at that rating (except for a couple more that we’re going to see soon!)
Baltimore starts off strong, becoming a very good squad before the 20s hit, and they hit hard. Baltimore struggles and becomes a team quite unused to success until S32 when the team starts its move back up. It’s hard fought, but with a dramatic leap at the end of S39, we have a Baltimore squad that is Ultimus caliber and remains so.
Yellowknife is a team that goes through the competition -> trough -> competition cycle that we see so much in the ASFC, it’s just that they’re better than those ASFC teams. The Yellowknife teams, even when in a rebuild, are always among the stronger teams in the league, with ratings that put them squarely in Ultimus territory, and no wonder, they’ve got among the most appearances in the league! If we project forward from their history, we’re going to see a strong 6 seasons where they lose in the finals a couple more times.
Cape Town (Philly Liberty) joins in S2, they start off strong, mirroring Yellowknife, but when Yellowknife rebuilds successfully, the Liberty continue to drop. They hit lows in the 30s before a rebrand and a team that successfully builds itself back up as a team in contention.
Chicago joins the league on a high before utterly dropping down the ranks. Chicago slowly but surely builds itself up, remember that they are about as well rated as the average ASFC squad at this point, and peak in the early 40s before another tumble. They are on the upswing, but with the age of the squad, I’m not sure how this chart will go.
Sarasota is a team used to success and has stayed at above 1500 for most of their lifespan. The team has basically always been strong, although they face a consistent decline in the late 30s continuing til now. We’ll have to see how they do, as they don’t have a history of rebuilding.
Berlin shares an arc that’s very close to the old Legion into New Orleans swap, a strong start of improvement from expansion that competes before just kinda dropping away and not reaching those heights again. With their rebuild going, we’ll see what their outlook is like.
We can also pull up a chart of all our teams currently, we’ve got a quick and easy look at the general history of the teams here.
However, this table doesn’t show us much, especially compared to the charts, so we’re going to look at some better numbers, some more cool ones. We’re not super concerned with teams today, we’re interested in the history of the teams with this article.
Although, with the table, we can see some fun numbers.
First off, I notice a lot of green at the bottom of the table in the Min/Avg/Median section. Orange County, Sarasota, and Yellowknife are just habitual winners, with Arizona joining them.
Colorado has had the most tragic season with a 98 point drop, and we probably won’t see that again, that’s a quirk of early season data and model uncertainty.
New York has a surprisingly even-keeled set of seasons, never really dropping very far if at all. And lastly, Baltimore with a huge 108 point gain over a single season. Similar to the Colorado drop, this is mostly model uncertainty in early seasons.
Alright, now that we’ve seen some history, let's look at the top teams of all time!
Hm, well, unsurprisingly the team with multis that won three times in a row during model calibration (high gains/losses) peaks very highly. Let’s remove those seasons from the data set and see what we get.
Well, this is also pretty unsurprising. The ISFL meets the start of the incredibly long Otters run, but Chicago sneaks in there too with their brilliant first season!
How about we remove those OCO seasons like we did the AZ seasons and what happens?
Well, I don’t know what else to expect there. Let’s just remove Arizona and Orange County entirely.
Finally! Some more teams! The early Chicago teams were fantastic and since a team’s earliest seasons are the most volatile, they have some highly rated teams (a franchise who would not do so well for the next while.)
We also see the existence of the Baltimore and Yellowknife duo; these two ran the conference until Baltimore’s fall and Sarasota’s rise.
Conversely, we’ve also got the bottom teams. Austin gets unfairly dinged for their first seasons with model uncertainty combined with losing, but we’ve also got early Las Vegas there as they try to dig their way out of that hole.
Beyond that, this kind of gave me an idea, what if we look to see if we can’t find “dynastic” runs, or beyond that, very highly peaking teams?
If an Ultimus winner is around 1540 rating, a top top squad should be around… 1575?
So, I guess this was to be expected. We’ve got a couple colour coded runs for the Outlaws, the early seasons (we already know why there’s an asterisk here) before really rising in the late 30s. Chicago has their first season before dropping down (model uncertainty) and then we have Orange County basically during the entire duration of the First Sim Era. Yellowknife also peaks as a Dominant Team in S8, a season where they won an Ultiimus.
The “Un-Dominance Finder” looks for teams below 1375 rating, and we get a lot more hits, although the entire Austin, Berlin, Honolulu, New York, and San Jose ones are just there because of model uncertainty.
Baltimore had a truly tragic stretch Starting around S25 that bottomed out during 29-32.
Chicago had a pair of bad seasons in 23-24 as they just tumbled from their first 10-3 season.
Colorado had a really rough go of it. I marked this as two troughs, a really bad start into a resurgence into a rebuild will do that.
New Orleans really did badly as the Legion before digging themselves out, and of course we have the infamous tank that they are similarly pulling themselves out of.
San Jose has a single season that isn’t an early season drop, but a single season that went badly isn’t anything to talk about really.
Alright, we looked at some historical data some more, but what about the Ultimus? Which is the best and worst squad to win an Ultimus? How about the biggest upset?
First off, the first Ultimus Bowls are wracked by model uncertainty and so they don’t count very much (also the asterisks lol).
But when you look at it, barring a few major upsets, generally speaking, the better team won. We also have a surprisingly large rating differential between the winner and loser, very often.
Most Ultimus teams are really good. The worst team to win Was Baltimore in S33, but that’s a team that’s coming up from a really really bad stretch. These ratings don’t tell you how well a team is performing recently as they carry the baggage from their previous seasons.
One number that does stand out, however, is the Largest Differential. S8, was by rating, the greatest steam-roll one could have expected, the difference between Yellowknife and New Orleans was a staggering 257 points, basically the difference between the worst and best team every season.
The biggest upset was in S22 when a very overmatched Colorado went into Orange County and came out victorious.
This also kind of begs the question of sim luck. This one is really hard to nail down, as I can’t really figure out a way to properly display it, but how well do teams perform in the Ultimus compared to their opponents?
Austin has generally brought a weaker squad to the final game than their opponent, but have a differential of 0 (2 wins and 2 losses) so we can say that they’re sim blessed (probably).
Arizona is better than their opponents by the same margin that Austin is rated weaker, but are a +7 for 10 total wins, is that sim-blessed or just performing to expectation?
I think generally speaking, we can point out a couple teams that haven’t done well, or did well. Austin, from before, probably over-performed. Similarly, Sarasota underperformed their rating, with a losing differential but bringing a stronger side than their opponents in the Ultimus. Yellowknife is the same, their strong sides have not won as much as they should, and their weak sides have lost a little more than they should. San Jose is the biggest sim-blessed side as their wins have come with quite overmatched squads!
I had a lot of fun figuring out this Glicko/web-scraping stuff, just something cool I’ve never done before and so this article is just a bonus. Was cool seeing all this data and seeing the charts. The spreadsheet is linked above, and the data info is down below.
Data Info - lemonoppy ISFL Game API
You can find a list of all games there (S1-47) and there’s some query params you can do to narrow down the data you get returned. Basically, you add a parameter to the URL and then you get a more specific set of data.
Specific Team: https://www.lemonoppy.com/api/isfl/games?team=AZ
Specific Season https://www.lemonoppy.com/api/isfl/games?season=47
Specific Week: https://www.lemonoppy.com/api/isfl/games?week=13
Winner: https://www.lemonoppy.com/api/isfl/games?winner=AZ
Loser: https://www.lemonoppy.com/api/isfl/games?loser=HON
Playoff Games only: https://www.lemonoppy.com/api/isfl/games?playoffs=true
Home Team: https://www.lemonoppy.com/api/isfl/games?home=AZ
Away Team: https://www.lemonoppy.com/api/isfl/games?away=AZ
And you can combine all of these for a fun query as well, for example: https://www.lemonoppy.com/api/isfl/games...&loser=HON would get you a list of all the games where Arizona beat Honolulu in the playoffs (S37, S41, S42, S43)
or something like: https://www.lemonoppy.com/api/isfl/games...JS&away=AZ is all the time AZ has beat SJS in SJS
The shape of a game has the teams involved, when/where the game happened, and the scores of each team in each of the quarters in the game. (Overtime defaults to -1 if it didn’t happen).
{
"season": 43,
"week": 100,
"away_team": "HON",
"home_team": "AZ",
"away_score": 27,
"home_score": 32,
"winner": "AZ",
"loser": "HON",
"overtime": false,
"away_first": 0,
"away_second": 21,
"away_third": 3,
"away_fourth": 3,
"away_overtime": -1,
"home_first": 10,
"home_second": 7,
"home_third": 8,
"home_fourth": 7,
"home_overtime": -1
}
Notes:
Pre-season:
-4 = Preseason Week 1
-3 = Preseason Week 2
-2 = Preaseson Week 3
-1 = Preseason Week 4
Playoffs:
100 = Wildcard Game
200 = Conference Champs
300 = Ultimus
First, a little technical bit on how I got the data. I recently started scraping the ISFL Index for Win/Loss data. Initially I was just doing it to see what scraping was like, I never really do it and so it was a fun project and I thought I could make an Elo thing but historical. I found that it was pretty easy to do, and that led me to having a data set (a big JSON file) of every single game that has been played in ISFL history (the ones found here: https://index.sim-football.com/ISFLS47/GameResults.html) and I thought that that was neat.
I was kinda annoyed with the shape of the data so I ended up just putting it all in a database and made a “dynamic” searchable API for anyone who might want the data for media of their own. It’s a fun little thing and I’ll add details at the end for anyone who wants to play with it. Let me know if there’s any other data scraping you would be interested in, I’m just going to build a big data thing so people can write media easier or satisfy their curiosity.
Anyway, with all this data, I realized I could do a historical Elo calculation of all teams. Elo is a rating system that tells you how good a team is based on how you win/lose to other teams. However, there’s a “better” system called Glicko (and Glicko2 as an improvement on that). Took me an hour to write the code for the info we have today.
What a Glicko rating is is a rating based on a set of matches, ie. 1 season. A team’s rating goes up or down based on who they beat within a season, but Glicko also has a built in mechanism for uncertainty on your rating if your team is winning or losing in an erratic manner (ie. Yellowknife these past seasons), or if your team is new to the dataset (the first 3-4 seasons of a team are kinda inaccurate because of model uncertainty).
TL;DR - Rating high = team good. First three seasons of a team's history are less accurate.
I used this data I had gotten programmatically to map the ratings of teams, and threw it all into a spreadsheet where we can make some cool charts and tables!
(https://docs.google.com/spreadsheets/d/1...sp=sharing)
Historical ISFL Team Rating
Alright, first up are the big charts!Everyone loves charts! This is the history of the ISFL by rating. The average rating of the league is 1470.68 through all the seasons, and the median is about 1467. Notably, most teams in the league live around 1520-1410ish, and teams that go above or below that are pretty exceptional in one way or another. You’ll also see big swings the younger a team is, as the model tries to narrow down their certainty of the team’s performances.
Some key numbers:
So, if you’re a team competing for a championship, you want a rating of about +50 on the median, and conversely a bad team is about -50 on the median.
That chart above has way too much going on, so I’ve just copy and pasted the chart but only showing the ASFC teams. We’ll just go through each conference separately so we can actually see what’s happening.
The ASFC is characterized, generally, by a set of middling teams and then Orange County and Arizona standing way above them. This is going to be more apparent the further we go in this piece.
Arizona starts off the league strong with a three-peat under some “mysterious” circumstances (not mysterious, it was some “mild cheating”). After the multis get busted, Arizona becomes about an average ASFC team (which is a bad NSFC team) and starts improving around the Current Sim Era where they’ve reached pretty dominant success.
That Arizona dominance is really nothing compared to Orange County. The Otters are pretty unparalleled in the ISFL. I think the chart scale really does a disservice to how disgustingly good the Otters were for the first half of the league. Remember, the median league winner is a 1543 rating, the Otters went from S4-26 at 1560 rating, and were comfortably at 1600 for a long while. Of course, starting S27, they’ve slumped and have not had positive seasons since.
San Jose is our third Original 6 team, we’ve got some spikes at the fore-front of their history but they have only peaked at 1500+ once (S6) before being a 1440ish team, which is below the league median.
New Orleans (or the Las Vegas Legion) are our next team to join the ASFC. The Legion started out ROUGH. However, in S7, their fortunes started to change and they rocketed up the standings before turning into a league-average team and then starting to drop with the Current Sim Era and then of course the 40s started another huge down-tick. This past season however, sees a big rise, and we’re going to see them shoot up again as their low rating combined with wins will rocket them up.
Austin similarly started the league in a very rough space. No problem for this expansion team though, as they quickly built themselves as a competitor in the early 30s and managed to peak in the 40s as a championship team.
Honolulu joined the league in S22 and had a fantastic run in the mid 30s, clearly separating themselves from the rest of the ASFC before a tank, another rise, and then a sharp drop in the current seasons.
New York follows a trajectory that is pretty familiar by now, a rough start for an expansion team before quickly becoming a top squad for a few seasons, a partial rebuild trough, and then competition. This really is the ASFC, way more than the NSFC that we’ll see soon.
The NSFC is markedly different from the ASFC. First off, we don’t have that consistent 1600 rating squad that we saw, and also, the median is just higher. The NSFC teams, as a whole, are more consistent than the ASFC. The ASFC is marked by teams being really good or bad, and the NSFC has more parity within it.
Starting with Colorado, we see a team that struggled mightily at the start, coming to a nadir in S7, before a hard-fought battle before a strong S25-35 before dropping back down. Colorado is a pretty unique case, you don’t often see 10 season stretches at that rating (except for a couple more that we’re going to see soon!)
Baltimore starts off strong, becoming a very good squad before the 20s hit, and they hit hard. Baltimore struggles and becomes a team quite unused to success until S32 when the team starts its move back up. It’s hard fought, but with a dramatic leap at the end of S39, we have a Baltimore squad that is Ultimus caliber and remains so.
Yellowknife is a team that goes through the competition -> trough -> competition cycle that we see so much in the ASFC, it’s just that they’re better than those ASFC teams. The Yellowknife teams, even when in a rebuild, are always among the stronger teams in the league, with ratings that put them squarely in Ultimus territory, and no wonder, they’ve got among the most appearances in the league! If we project forward from their history, we’re going to see a strong 6 seasons where they lose in the finals a couple more times.
Cape Town (Philly Liberty) joins in S2, they start off strong, mirroring Yellowknife, but when Yellowknife rebuilds successfully, the Liberty continue to drop. They hit lows in the 30s before a rebrand and a team that successfully builds itself back up as a team in contention.
Chicago joins the league on a high before utterly dropping down the ranks. Chicago slowly but surely builds itself up, remember that they are about as well rated as the average ASFC squad at this point, and peak in the early 40s before another tumble. They are on the upswing, but with the age of the squad, I’m not sure how this chart will go.
Sarasota is a team used to success and has stayed at above 1500 for most of their lifespan. The team has basically always been strong, although they face a consistent decline in the late 30s continuing til now. We’ll have to see how they do, as they don’t have a history of rebuilding.
Berlin shares an arc that’s very close to the old Legion into New Orleans swap, a strong start of improvement from expansion that competes before just kinda dropping away and not reaching those heights again. With their rebuild going, we’ll see what their outlook is like.
Team Tables
We can also pull up a chart of all our teams currently, we’ve got a quick and easy look at the general history of the teams here.
However, this table doesn’t show us much, especially compared to the charts, so we’re going to look at some better numbers, some more cool ones. We’re not super concerned with teams today, we’re interested in the history of the teams with this article.
Although, with the table, we can see some fun numbers.
First off, I notice a lot of green at the bottom of the table in the Min/Avg/Median section. Orange County, Sarasota, and Yellowknife are just habitual winners, with Arizona joining them.
Colorado has had the most tragic season with a 98 point drop, and we probably won’t see that again, that’s a quirk of early season data and model uncertainty.
New York has a surprisingly even-keeled set of seasons, never really dropping very far if at all. And lastly, Baltimore with a huge 108 point gain over a single season. Similar to the Colorado drop, this is mostly model uncertainty in early seasons.
Top Teams of all Time
Alright, now that we’ve seen some history, let's look at the top teams of all time!
Hm, well, unsurprisingly the team with multis that won three times in a row during model calibration (high gains/losses) peaks very highly. Let’s remove those seasons from the data set and see what we get.
Well, this is also pretty unsurprising. The ISFL meets the start of the incredibly long Otters run, but Chicago sneaks in there too with their brilliant first season!
How about we remove those OCO seasons like we did the AZ seasons and what happens?
Well, I don’t know what else to expect there. Let’s just remove Arizona and Orange County entirely.
Finally! Some more teams! The early Chicago teams were fantastic and since a team’s earliest seasons are the most volatile, they have some highly rated teams (a franchise who would not do so well for the next while.)
We also see the existence of the Baltimore and Yellowknife duo; these two ran the conference until Baltimore’s fall and Sarasota’s rise.
Conversely, we’ve also got the bottom teams. Austin gets unfairly dinged for their first seasons with model uncertainty combined with losing, but we’ve also got early Las Vegas there as they try to dig their way out of that hole.
Beyond that, this kind of gave me an idea, what if we look to see if we can’t find “dynastic” runs, or beyond that, very highly peaking teams?
If an Ultimus winner is around 1540 rating, a top top squad should be around… 1575?
So, I guess this was to be expected. We’ve got a couple colour coded runs for the Outlaws, the early seasons (we already know why there’s an asterisk here) before really rising in the late 30s. Chicago has their first season before dropping down (model uncertainty) and then we have Orange County basically during the entire duration of the First Sim Era. Yellowknife also peaks as a Dominant Team in S8, a season where they won an Ultiimus.
The “Un-Dominance Finder” looks for teams below 1375 rating, and we get a lot more hits, although the entire Austin, Berlin, Honolulu, New York, and San Jose ones are just there because of model uncertainty.
Baltimore had a truly tragic stretch Starting around S25 that bottomed out during 29-32.
Chicago had a pair of bad seasons in 23-24 as they just tumbled from their first 10-3 season.
Colorado had a really rough go of it. I marked this as two troughs, a really bad start into a resurgence into a rebuild will do that.
New Orleans really did badly as the Legion before digging themselves out, and of course we have the infamous tank that they are similarly pulling themselves out of.
San Jose has a single season that isn’t an early season drop, but a single season that went badly isn’t anything to talk about really.
Ultimus and Ratings
Alright, we looked at some historical data some more, but what about the Ultimus? Which is the best and worst squad to win an Ultimus? How about the biggest upset?
First off, the first Ultimus Bowls are wracked by model uncertainty and so they don’t count very much (also the asterisks lol).
But when you look at it, barring a few major upsets, generally speaking, the better team won. We also have a surprisingly large rating differential between the winner and loser, very often.
Most Ultimus teams are really good. The worst team to win Was Baltimore in S33, but that’s a team that’s coming up from a really really bad stretch. These ratings don’t tell you how well a team is performing recently as they carry the baggage from their previous seasons.
One number that does stand out, however, is the Largest Differential. S8, was by rating, the greatest steam-roll one could have expected, the difference between Yellowknife and New Orleans was a staggering 257 points, basically the difference between the worst and best team every season.
The biggest upset was in S22 when a very overmatched Colorado went into Orange County and came out victorious.
This also kind of begs the question of sim luck. This one is really hard to nail down, as I can’t really figure out a way to properly display it, but how well do teams perform in the Ultimus compared to their opponents?
Austin has generally brought a weaker squad to the final game than their opponent, but have a differential of 0 (2 wins and 2 losses) so we can say that they’re sim blessed (probably).
Arizona is better than their opponents by the same margin that Austin is rated weaker, but are a +7 for 10 total wins, is that sim-blessed or just performing to expectation?
I think generally speaking, we can point out a couple teams that haven’t done well, or did well. Austin, from before, probably over-performed. Similarly, Sarasota underperformed their rating, with a losing differential but bringing a stronger side than their opponents in the Ultimus. Yellowknife is the same, their strong sides have not won as much as they should, and their weak sides have lost a little more than they should. San Jose is the biggest sim-blessed side as their wins have come with quite overmatched squads!
I had a lot of fun figuring out this Glicko/web-scraping stuff, just something cool I’ve never done before and so this article is just a bonus. Was cool seeing all this data and seeing the charts. The spreadsheet is linked above, and the data info is down below.
Data Info - lemonoppy ISFL Game API
You can find a list of all games there (S1-47) and there’s some query params you can do to narrow down the data you get returned. Basically, you add a parameter to the URL and then you get a more specific set of data.
Specific Team: https://www.lemonoppy.com/api/isfl/games?team=AZ
Specific Season https://www.lemonoppy.com/api/isfl/games?season=47
Specific Week: https://www.lemonoppy.com/api/isfl/games?week=13
Winner: https://www.lemonoppy.com/api/isfl/games?winner=AZ
Loser: https://www.lemonoppy.com/api/isfl/games?loser=HON
Playoff Games only: https://www.lemonoppy.com/api/isfl/games?playoffs=true
Home Team: https://www.lemonoppy.com/api/isfl/games?home=AZ
Away Team: https://www.lemonoppy.com/api/isfl/games?away=AZ
And you can combine all of these for a fun query as well, for example: https://www.lemonoppy.com/api/isfl/games...&loser=HON would get you a list of all the games where Arizona beat Honolulu in the playoffs (S37, S41, S42, S43)
or something like: https://www.lemonoppy.com/api/isfl/games...JS&away=AZ is all the time AZ has beat SJS in SJS
The shape of a game has the teams involved, when/where the game happened, and the scores of each team in each of the quarters in the game. (Overtime defaults to -1 if it didn’t happen).
{
"season": 43,
"week": 100,
"away_team": "HON",
"home_team": "AZ",
"away_score": 27,
"home_score": 32,
"winner": "AZ",
"loser": "HON",
"overtime": false,
"away_first": 0,
"away_second": 21,
"away_third": 3,
"away_fourth": 3,
"away_overtime": -1,
"home_first": 10,
"home_second": 7,
"home_third": 8,
"home_fourth": 7,
"home_overtime": -1
}
Notes:
Pre-season:
-4 = Preseason Week 1
-3 = Preseason Week 2
-2 = Preaseson Week 3
-1 = Preseason Week 4
Playoffs:
100 = Wildcard Game
200 = Conference Champs
300 = Ultimus
Ultimus: S46, S47
ISFL Most Dedicated Member: S46
Gemini Awards: S42 Best Article (Series), S44 Best Article (Limited), S46 Best Author, S47 Best Author, S48 Best Article (Limited)
DSFL Most Dedicated Member: S42
Getting Defensive Podcast: S42 New Player Silver Medalist
ISFL All-Rookie Team: WR1 (S45)
DSFL Offensive Player of the Year: S43
DSFL First-team All-Pro: S43, S44
DSFL Pro Bowl: S42, S43, S44
S43 R1.01 - Arizona Outlaws
S42 R1.04 - Minnesota Grey Ducks
-----
Try out the Player Updater tool!
lemonoppy.com/isfl/player-updater
lemonoppy.com/isfl/player-updater