#56 Recruiting Efficiency Score

[alexfall862]

Hello Coaches!

 

Welcome to the second Dev Diary for 2022 on the new recruiting system for this coming season.

 

Recruiting Dev Diaries for 2022

Dev Diary 54: Recruiting Affinities Refresh | Close to Home

Dev Diary 55: Recruiting Affinities Refresh | Other Affinities

Dev Diary 56: Recruiting Efficiency Score

Dev Diary 57: Recruiting on the Interface

 

With this diary, I want to introduce the new recruiting system. The goal of overhauling the system was to take advantage of what the interface gives us in the form of new ways to connect on-field performance and recruiting while both preserving and nerfing the affinity system.

 

Challenges of Last Season

 

One of the things that was clear when reviewing the first season of recruiting was that affinities had a very strong influence on whether a team could even consider competing for a recruit. Schools could invest as little as 5 points into a recruit and get 90% of their necessary points through the affinity bonus system. In isolation, this doesn't mean too much, but the effect it had on recruiting was in shutting out competition and leading to a dice roll in player generation dictating committals before recruiting even began.

 

We can also take advantage of the interface, and more realistically simulate the way a player may consider a school. Beyond affinity matching having a strong hand in recruiting choices, we also saw Georgia and Nebraska bring in really strong classes despite absolutely putrid seasons. To be fair, in real life poor teams can often out recruit their relative place in the standings due to affinity with recruits, but while both teams were closing the season getting blown out, they were also signing 5* recruits and that just didn't pass the smell test.

 

The new recruiting system retains the points based recruiting backbone of the previous system. Each team still gets 50 Recruiting Points (now called Recruiting Time Points - think of the 50 points as an approximation of how your coaching staff divides their time among recruits) each week. (or fewer if you have an administrative penalty for bad gameplans or other site issues).

 

Recruits still have a threshold that ticks down over time and reacts to the number of teams recruiting them.

 

What has changed are two things: we've swapped out the Affinity Bonus for a Recruiting Efficiency Score and recruits now have memories.

 

Recruits are Now Goldfish

 

Before we dive into the big change, let's go over memories. One of the bigger issues that came up in feedback was the battle of trying to overcome schools who were currently inactive but had pushed the committal threshold so high that no other school could get a player to commit until the last week of the recruiting season.

 

Part of this was intended. Recruits should naturally want to wait to hear back form all possible options - but when looking at some of the recruiting battles, it was clear that this behavior from recruits was ultimately harmful to the fun of recruiting and not all that realistic as a recruit is unlikely to commit to a school that's ghosting them.

 

Recruits will now only evaluate the last ten weeks when counting the number of teams interested in them. They will still consider all Recruiting Time Points when calculating a decision, though, so a school that was active in week 1 could conceivably get a commitment without further activity provided no other school made a serious effort on that player.

 

Recruiting Efficiency Score

 

The Recruiting Efficiency Score is a modifier to team's point totals with recruits meant to be an approximation of how well your pitch is landing with a recruit. 0-8 with no affinity match with a recruit? Expect your pitch to be less effective with a recruit than a 2-6 school with 1 affinity match, or a an 8-0 juggernaut that's also close to home.

 

The value of the modifier in the Recruiting Efficiency Score is between -20% and 20%, meaning that the worst possible fit for a recruit is at 80% strength compared to the median school. It also means that the best possible situation has effectively 120% strength in recruiting. When pitting the best fit against the worst fit, it would be a 50% advantage for the best fit school. That's far more modest than the previous affinity system gave out.

 

The nuts and bolts of how the affinity score is calculated are below, but broad strokes are this: wins and losses account for 30% of the score, postseason participation of various kinds and strengths (bowl game, conference championship appearance, conference title, playoff, etc.) account of 10%, and 50% of the score is evaluated by matching affinities.

 

 

 

The specific weights for the current calculations of the RES are as follows:

Item	Weights
Previous Season Overall Record Wins over Conference Games Played	5.00%
Previous Season Conference Wins over Conference Games Played	5.00%
Current Season Overall Record W over Games Played	10.00%
Current Season Conference W over Games Played	10.00%
Coach Career Record Wins (avg wins per season over total games per season)	10.00%
Affinity Match 1	25.00%
Affinity Match 2	25.00%
Postseason Games (Full Credit for Bowl Game Participation, Bowl Outcome has No Weight)	5.00%
Conference Championship (Full Credit for Title, Half Credit for Participation)	5.00%
Rivalry Trophies Held / Other Flavors (WIP, may be added for 2023)	0.00%

 

 

 

Examples

 

This is all a little heady, so let me give some example situations.

 

Tulsa had a perfect season last year. They maxed out every category in the RES that a school can control. Their score for the year was 100%.

 

If they're recruiting a player they have 1 affinity match for, their RES goes to 110%, and 120% if they match both affinities.

 

Because this is a modifier on the Recruiting Time Points, Tulsa's RES for a 1 affinity match recruit would be:

 

Week 1: (5 RTP Points in Week 1) * RES of 110% = 5.5 Points logged into the recruit in week 1.

 

Let's say Tulsa loses their opener. That would impact their score (0/0 games played is full credit in the RES, but once a game is played it references that outcome. 0/0=1 is meant to be an approximation of hope springing eternal in the offseason.) This would drop their baseline RES score from 100% to 94.76%. The affinity bonus still applies, so in week 2 their RES would be 104.76%.

 

Week 2: (5 RTP Points in Week 1 + 5 RTP Points in Week 2) * RES of 104.76% = 10.476 Points logged into the recruit week 2.

 

Oh no, they also lose their home conference opener as well! But, to make up for that, they're spending 10 RTP points this week on the recruit instead of 5. Their baseline RES drops to 89.56% because they're now getting 0 credit for current season conference record.

 

Week 3: (5 RTP Points in Week 1 + 5 RTP Points in Week 2 + 10 RTP Points in Week 3) * RES of 99.56% = 19.912 Points logged into the recruit week 2.

 

Competition

 

Let's now say that Texas Tech has an interested in that same player and also has an affinity match, but where Tulsa had the best season, Texas Tech has the worst. For comparison of the system, let's say they start off white hot and win their first game and conference opener.

 

Their initial RES is 90% and they have 1 affinity match for an RES with this recruit of 100%. Let's say they win a noncon week 1 and a conference game week 2 and go into week 3 sensing a battle with Tulsa and also put in 10 puts in week 3. This is how it would break down:

 

Week 1: (5 RTP Points in Week 1) * RES of 100% = 5 Points logged into the recruit in week 1.

 

Week 2: (5 RTP Points in Week 1 + 5 RTP Points in Week 2) * RES of 100.29% = 10.0029 Points logged into the recruit week 2.

 

Week 3: (5 RTP Points in Week 1 + 5 RTP Points in Week 2 + 10 RTP Points in Week 3) * RES of 100.53% = 20.106 Points logged into the recruit week 2.

 

So we can see that Texas Tech started the season at a 10% disadvantage, but in just 3 weeks were able to shorten the gap in RES score, by virtue

 

In a third example, let's say Texas Tech has all the plans in the world to contest this recruit, but instead of opening white hot, they fall flat on their face while Tulsa continues on pace.

 

Tulsa by Week 3: Baseline RES of 100% + 10% Affinity Bonus = 110% * 20 = 22 points logged.

Tech by Week 3: Baseline RES of 80% + 10% Affinity Bonus = 90% * 20 = 18 points logged.

 

In a fourth example, let's say everything about the previous example stays the same, except Tulsa has no affinity match with this particular recruit.

 

Tulsa by Week 3: Baseline RES of 100% + 10% Affinity Bonus = 100% * 20 = 20 points logged.

Tech by Week 3: Baseline RES of 80% + 10% Affinity Bonus = 90% * 20 = 18 points logged.

 

Takeaways

 

The example above uses the best and worst team possible by on-field performance. Tulsa currently has the highest possible score, and Texas Tech has the lowest possible score. For most teams, their baseline RES scores are going to revolve around 90-95% throughout the season. While that's not nothing, it isn't a huge difference in impact on the total race for a recruit.

 

However, cumulatively, a bad season is about a 10-15% difference in RES and it takes roughly 1 affinity match to counteract that deficit. This has the intended goal of making teams that should have no business in real life of competing for recruits against powerhouse programs think twice before chasing after 5* players - yet at the same time doesn't completely eliminate the possibility they could get one.

 

Conference championships and bowl games last until the next round of games, so they can act as a buoy for teams that have a bad follow-up year. This advantage doesn't seem unrealistic, as the 'glow' doesn't come off a team in less than a year.

 

The longer the sim goes, the more stable a coaching record value is going to be. This should act as a stabilizing force for many programs with long-tenured programs, and will also protect coaches who take over struggling teams from being tied behind the recruiting anchor of their new team's previous season record. A 9-3 good coach taking over a 3-9 program equals out to the same credit as a 6-6 team with a 6-6 coach.

 

2021 End of Year Scores

 

The RES score is supposed to be an evolving number, but to give you some idea of of the effect it will have by year's end, here's what last year's end of season RES would have looked like:

 

Without the bump from the RES calculator treating the beginning of the season as a 'clean slate', Texas Tech sits at 80% along with UTEP.

 

 

 

 

Team	SUM of Recruiting Efficiency Score
Tulsa	100.00%
Florida State	99.31%
Kentucky	99.27%
Wisconsin	98.53%
Texas State	98.43%
Ohio	97.02%
North Texas	97.02%
Boston College	96.43%
Old Dominion	95.59%
Alabama	95.59%
Baylor	95.46%
San Diego State	95.30%
Purdue	95.09%
Illinois	95.09%
Texas	95.04%
Air Force	94.80%
Tennessee	94.71%
TAMU	94.71%
Oregon	94.67%
Arizona	94.26%
Florida Atlantic	93.87%
Akron	93.87%
Appalachian State	93.52%
California	93.28%
Virginia	93.24%
Temple	93.24%
UAB	93.02%
Tulane	92.99%
Troy	92.99%
Missouri	92.99%
Mississippi State	92.99%
Marshall	92.99%
Duke	92.99%
Cincinnati	92.99%
TCU	92.44%
Kansas State	92.44%
Colorado	92.44%
Western Michigan	92.14%
UNLV	92.14%
Central Michigan	92.14%
Stanford	91.66%
Oregon State	91.66%
Ohio State	91.66%
Michigan State	91.66%
Wyoming	91.30%
Utah State	91.30%
Rice	91.30%
New Mexico	91.30%
Bowling Green	91.30%
Arkansas State	91.30%
NC State	91.27%
Georgia State	91.27%
Georgia Tech	90.45%
UTSA	90.42%
SMU	90.42%
Pittsburgh	90.42%
Louisiana Tech	90.42%
Fresno State	90.42%
Auburn	90.42%
UCLA	89.97%
Oklahoma State	89.97%
Rutgers	89.71%
Coastal Carolina	89.58%
Penn State	89.19%
Minnesota	89.19%
Maryland	89.19%
Ball State	89.12%
East Carolina	88.96%
Nevada	88.08%
Navy	88.08%
Lousiville	88.08%
Florida International	88.08%
Florida	88.08%
USC	87.69%
Kansas	87.69%
Iowa State	87.69%
Arizona State	87.69%
Northern Illinois	87.21%
North Carolina	87.21%
Middle Tennessee	87.21%
Kent State	87.21%
Eastern Michigan	87.21%
Colorado State	87.21%
Birginia Tech	87.21%
Arkansas	87.21%
BYU	86.92%
Michigan	86.92%
West Virginia	86.78%
Northwestern	86.78%
UCF	86.33%
South Carolina	86.33%
Miami (OH)	86.33%
Georgia Southern	86.33%
Ole Miss	86.29%
Louisiana	86.29%
Iowa	86.14%
Umass	86.08%
San Jose State	85.42%
Memphis	85.42%
Buffalo	85.38%
Washington	85.22%
Oklahoma	85.22%
Indiana	85.22%
Notre Dame	84.58%
Army	84.58%
Western Kentucky	84.50%
Wake Forest	84.50%
USF	84.50%
Southern Miss	84.50%
South Alabama	84.50%
Miami (FL)	84.50%
Clemson	84.50%
Washington State	84.31%
UConn	83.67%
New Mexico State	83.67%
LSU	83.63%
Nebraska	83.39%
Liberty	82.75%
Toledo	82.71%
Syracuse	82.71%
Louisiana-Monroe	82.71%
Houston	82.71%
Hawaii	82.71%
Charlotte	82.71%
Boise State	82.71%
Utah	82.61%
Vanderbilt	81.79%
Georgia	81.79%
UTEP	80.00%
Texas Tech	80.00%

 

 

 

Here's that distribution as a graph

 

 

 

 

 

 

 

and the quartile measurements - lower scores tend to be lower than the high scores are high, meaning that having a bad season is going to separate you from the meh teams more than than a meh season is going to separate you from the good teams. Again, this is the intended outcome. That's not to say teams that perform poorly can't recruit good players, they'll just need to be more mindful of affinity matches and spend more points per recruit.

 

 

 

 

Min	80.00%
2nd Quartile	85.58%
3rd Quartile	89.19%
4th Quartile	92.99%
Max	100.00%

 

 

 

 

 

 

This is a WIP

 

This project is a WIP. Other derivations of this general idea are also possible.

 

Testing on this concept has been slow, because the number crunching is just a little too much for a google sheet to handle, and I'm sure we'll learn things when we do a test run off of the interface. As stated above, the goal here is to take advantage of what the interface gives us in the form of new ways to connect on-field performance and recruiting while both preserving and nerfing the affinity system.

 

I could foresee the overal span of effect changing from the 80%-120% to 70%-130% or 90%-110% if coaches feel on-field performance has too dynamic an effect or not dynamic enough of an effect.

 

 

Wrap Up

 

Unlike the affinity refresh project, this is more of a single person project - so I'll take credit for whether this sinks or swims. I would like to shout out @TuscanSota though for getting the interface closer and closer to up and running and for talking me through what he can do the facilitate this idea. As I mentioned above, this is a WIP, but thanks to @TuscanSota, the bones of this project are ready to go in the interface.

 

Like everything else, I welcome your input and these diaries are an effort to be as transparent as possible about the way recruiting works, where my minds going as far as these projects, and I look forward to a test run in the near future. The next dev diary in this quartet will be written and released shortly as a preamble to the first test run of this recruiting system.

Edited April 10, 2022 by alexfall862

[rcarlson77]

"Texas Tech has the worst. For comparison of the system, let's say they start off white hot and win their first game and conference opener."

 

Can we make that white hot start thing happen?

[alexfall862]

Well Lubbock is pretty white and also it's hot there...

[rabidsnowman]

The value of the modifier in the Recruiting Efficiency Score is between -20% and 20%, meaning that the worst possible fit for a recruit is at 80% strength compared to the median school. It also means that the best possible situation has effectively 120% strength in recruiting. When pitting the best fit against the worst fit, it would be a 50% advantage for the best fit school.

 

So you use the term "median" for a team with 100 base score, then immediately only give that base score to Tulsa. IDK, they don't seem "median" to me. Feels like that throws off the whole +20/-20 scheme. Median looks a bit more like 90 to me. So +30/-10 feels a bit more accurate. I don't think this is a problem from a gameplay standpoint, and is actually probably more realistic than an even split anyway. I think the difference in the eyes of recruits between a 2-10 team and a 4-8 team is nowhere near the difference between a 8-4 and 10-2.

 

Edit: And you noticed this in your quartile numbers later too, so I'm guessing you didn't mean to call 100 the median school but rather the median score. Or I'm just dumb. IDK.

[alexfall862]

Oh sorry, the other 20 points are the affinities. So a perfect school would get 50% of the total possible score addition from that, and then another 20% comes from having or not having affinities. Broadly speaking, the difference between a 0-12 school and a 12-0 school is about the same as 1 affinity matching.

[Sarge]

I still don't know if I agree with wins and losses affecting recruiting until after sim year 4. Myself and other coaches are essentially recruiting with an additional handicap for a roster that was generated. I definitely think any win/loss bonuses should happen after 4 years of player recruiting. Otherwise, everything looks cool and exciting.

[smackemz]

I agree with @Sarge - these changes look awesome and exciting, but I also believe that any recruiting advantage based on performance should wait until season 4. I don't disagree that down the road performance and wins should impact recruiting, but seemingly teams that were blessed with strong generated rosters won a lot of games. They will now be blessed with recruiting advantages as well? Especially in what was considered a beta season.

 

Either way I'll play and have fun, but I think if we had a few seasons of truly equal footing recruiting it would be better long term.

[acewulf]

This is a very interesting system that is a very welcome addition to try and add in on the field accomplishments to the recruiting formula. I like it a lot.

 

I would also agree that the full extent of this doesn't need to be added until we have a more established basis for each of us. I think there are ways to incorporate this in gradually to where we can start to test this system out, but not have it be an insane advantage due to the state of the initial teams.

 

And if I am not mistaken, this is in addition to the affinities correct?

[BabaYaga]

I agree as well. I was really good last year, by mostly solely luck of my team draw.

 

This year I am probably going to be very bad because the underclass of my team was putrid.

 

I don't think Y1 or Y2 performance should affect recruiting but wait until we all get "our" players into our systems.