*S50 Fantasy - ADP, My Guys, and Ranking Rankers

[lemonoppy]

Fantasy football in the ISFL is kind of a weird concept, we’re doing fantasy football on fantasy players? And some of these people are also drafting Dynasty Fantasy teams as well? Just layers upon layers of fake players being super fun to watch and play. Well, if there’s a fantasy community, shouldn’t there be a fantasy analyst community as well? And then on top of that, analysis on analysis just to get even more into the weeds. Well, if that’s what you’re looking for, that’s what I’ve got for you!

Austin has done a ton of great work on the ADP of players and ranking the rankers, and I’ve got something similar for us today. I’ve got some ADP of players, ADP of ISFL teams, finding out the users/groups that strayed the furthest from ADP, and then finally we’ll see who you should believe in terms of ranking fantasy players. My media is a bit different in scope and approach but I hope it’s a fun read!

Points by ADP

This picture is part of my basis of the article, specifically that we’re going to be focusing on offensive positions only. The first defensive player taken in all of the 28 drafts is Wildboyy taking Crasher Wake (LB) 15th overall. Crasher Wake finished with 210 points, coming in below the top 10 LBs who ranged from 234 (Crazy Tomato NYS) to 211 in tenth (Joey Battle OSK). Simply put, the defensive positions are both not highly valued, they also don’t have a ton of “spikage”, the good players are good but not outstanding compared to their peers. We’re also going to be basically ignoring kickers, for the two of you who care.

Note: The 15th player in the WR/TE category by ADP was Seer Ial who finished as WR 6 and has much more of an interesting path to being a relevant fantasy pick from a pick value point of view.

Aside from only looking at the offensive players, we’re mostly going to be looking at the top 4ish rounds because that’s where our equity in analysis lies. After Round 4 and other players like defensive players start getting into the mix, we start losing any real edge, marginal players are marginal players and unless you can find a standout sleeper, we’re mostly looking at the data being skewed by filling in flex spots and QBs muddying the data.

Points by ADP comparing Flex-able Players vs Everyone

Note: That huge spike at 47 is Nick Williams, who was taken only once in the draft by @juniped, aka the GM of the player who got top of first round value for essentially nothing.

Teams and ADP

It makes sense that you want good players in NFL fantasy, JAGs or bad players aren’t likely to touch the ball or see the field, but the ISFL is different, we want touches and we want teams that are going to give players touches, wins or losses be damned because we’re probably not seeing any big roster changes mid-season. So, how do we determine who we’re going to take?

Loads of Team Data

This is a ton of data, it counts players, TPE, points per player, and points per player available in the Flex spot (RB/WR/TE) which we mostly will be focusing on. Let’s try and make some sense of it!

First off, before anything, the things I suspect will have any validity to them are all colour coded so that I can read them at a glance quicker, and those main ones are Flex Points/Player. We’re looking for the standout teams where we can get a ton of points and get an edge over the competition. Initially I thought that either a really bad or really good team is where you want to be, if your defense is really bad, that means your offense is probably going to be on the field more often after a quick score for example. This is actually not the case, there is basically no relation between how good or bad your offense is and how many points your flex player is going to score, not on the surface level.

TPE is an indicator of team “level” but more than raw TPE, the question for picking a player is a question of volume so we need to figure out some other way to filter to find fantasy stars.

Flex-able players by team, points, median ADP

Maybe if we look at points per unique player in the spot picked and compare their median ADP to their points?

It’s kind of a closer look. We find that players that get taken earlier are probably going to be worth more. Sometimes you have breakout players on bad teams like Orange County so you can find some big value as they just kind of do whatever they want to strat wise and pick up volume. But median ADP is often taken as a function of the ranker ADPs, so there’s some more questions there.

Teams sorted by season finish and seasonal Over/Unders

Alright, what if we go by seasonal O/U and actual seasonal Ws? This does look better!

From a first look, it looks like ADP is generally predictive of the performance of the fantasy player in question. The rankers generally do well at predicting players’ usage in the sim. However, we run into some important distinctions.

Rankers are good at predicting WR/TE ADP but much less good at predicting RB points. Running back points are generally in order of performance of the team. The better the team, the better the running back (which intuitively makes sense with winning teams running the ball to run out time, but is also a chicken/egg question).

WRs and TEs are predictable in usage based on the team usage and target spread to TPE distribution in the WRs. We’re often looking at stronger predictive WR1 usage and we might be getting some reach for WRs in ambiguous WR rooms where we kind of have to just hope we pick the right person. Figuring out the depth charts that teams will submit is very much more important than figuring out if they’re going to win for this spot.

I’ve got some guidelines on taking RBs in fantasy based off of this, and also the Otter Exception.

Otter Exception: You should take the top TPE ranked player in the RB or WR position on the worst team in the league but no others on their roster.
Note: This could also be called the Pitter Patter rule.

Running Backs should be taken if they’re on a good team, otherwise you’re probably not going to feel happy about it.

Wide Receivers and Tight Ends are great to pick based off of the Ranker ADP, they’ve done legwork on figuring out who’s going to be seeing the target share.

Ranking Rankers

Of course, a lot of this data is muddy because people are often coming from the same data source (the Fantasy Rankings) so there’s a lot of cross talk and insular noise rebounding, so, let's find who we should really listen to!

For these rankings, I’ve ordered players in the group and then created tiers if a set of players are close in points. For the ranking differential, I’ve treated the players in a tier as a block, so when Tuba ranked ranked Bugs as 12th but Bugs lives in the tier that spans from 5-7, Tuba gets a -5 since that is the difference from 12 to 5, not from 12 to 4 because the amount of points Bugs earned puts him as belonging anywhere between 5-7.

Rankers will also get dinged if they didn’t rank a player and they ended up in a tier that contains the #10 ranked player in the group, the penalty is 15 minus the best position.

QBs

First off, the Quarterbacks! Of the 14 QBs in the league, we had 12 taken, with Lloyd Bannings and Woflie McDummy Jr not being selected by a single player. Great for Bannings takers as they found someone who was at the replacement level at QB if they wanted him for free. The rankers really hit on the top QBs though, with McRobinson, Jimmies, Dyson, and Thor all hitting at the top of the range. Big L for br0_0ker as they ranked Thor at 12 but he ended up 4th.

The top guys seem pretty predictable and you can trust any of the rankers for the top 3 but with a huge caveat on the br0_0ker Thor hate.

RBs

Running Backs are the backbone of ISFL fantasy, there aren’t many true workhorses and you really need to hit on at least two running backs to win most groups, although the effect is lessened with the PPR bonus to WRs and increased bonus to TEs. RBs I counted rankings down to 15 for my scoring.

I think the most important thing here is the top hits. I want to be sure that a RB I take early is going to hit. Tuba takes the case here easily, his ranking differential was only 8 for his top 5 RB picks, with the big takeaways being the miss on Swaggert who underperformed pundit expectations and Shrubbery crushing it last season, who was a common miss.

Otherwise with top 5 picks, we’ve got similar miss rates, but Willeh really killed it with all the midrange RBs, slotting them into the tiers that they most likely ended up in, overcoming the Top 5 deficit to beat Tuba in the Top 10 category. We had a few big misses that added to score like the Hercules Henry, Maple Dogwood, and Kyle Crane underperformance that was shared amongst all the rankers.

Some big breakouts for this category included William Strong, Fig Newton, and Julian McMorris who were all ranked outside the top 15 or not at all but were amazing pickups.

WR/TEs

Wide Receivers and Tight Ends live in the same space so they’re ranked together. What we see is that most of the top 15 ranked players in this section did really well! Basically everyone got put where they were ranked, and accuracy reigned. But, if you look at the Top 10 delta section, you’ll see huge numbers across the board, what happened?

Two players tanked our rankers here, with La’Fluke Paris-Johnson and Sem’ga Nah’sim busting hard or swapping positions, we had a huge hole in our rankings. Once you adjust for that, you have this as the best position group to trust our rankers on, they’re super accurate and basically don’t miss. Tuba is the best by a lot in the Top 5 picks, and Top 10 while br0_0ker is your second choice for Top 10 ranked players in the group.

On the positive side, no one ranked @CROney3 into their top 15 but Benji put up a fantastic season, ending up as WR4, which also dinged all the drafters to the tune of 11 points.

Rankers Overview

At the end of it all, we’ve got a tier list to make for our rankers. I think that we can see that Tuba is our best ranker for the RB and WR/TE spots, but with omg and Willeh coming in to score big with their QB hits, I might want to take advice from them from that spot. Br0_0ker performs okay, they do better than I would do, but you’re probably better off using them as an option to check your leftover WRs than a first resource.

Ranking Groups on ADP

Now that we know who our best rankers are, let’s figure out our top My Guys and Ranking Bible Truthers in the ISFL!

If we look at users by adding up the pick they got a player at compared to their ADP, we can figure out who was the one to just take people in the tier list as gospel, and who took players just based on vibes.

We’ve got the top 10 in both categories here and the difference is… stark. The Rankers are the Rankers for a reason, all of the top 10 in ADP adherence found great value and finished in the points, while the most common finish for the My Guys players was 6th, except if you’re goated and negs and finish the top of the group while doing your own thing.

Also, as an aside, do you remember how @juniped got a free 1OA pick value because of Nick Williams?

We’re going to see some similar stuff in the biggest reaches and values, but it’s funny to see who people really believed in and thought were bums!

Conclusion

Do I have any advice for fantasy? Mostly it’s to trust the rankers on WR/TE, take RBs on good teams early, and if you’ve got a team like Orange County or Berlin a few seasons ago, their top player at the RB and/or WR spot are probably going to get a million touches because they’ve got no one else!


PS. Oh, the best QB/Flex core you could have gotten that no one else drafted and the spot they belong based on tiers was:

QB - Lloyd Bannings (QB 5)
RB - Nakiri Ayame (RB 23)
WR/TE - Big McLarge Huge (WR/TE 22)
WR/TE - Soul King Brook (WR/TE 22)
WR/TE - Kurt Wagner (WR/TE 22)

This team probably beats Juni's team!

[juniped]

> This team probably beats Juni's team!

Aw man that was unecessary!