Why Talent/Recruiting Rankings Matter

GopherWeatherGuy

Well-known member
Joined
Oct 24, 2013
Messages
15,354
Reaction score
13,668
Points
113
Since there are still several people on this site that don't believe in recruiting rankings, or that talent matters, I'll try to make it even more simple to understand.

247 has a yearly team talent composite dating back to 2015. https://247sports.com/Season/2017-Football/CollegeTeamTalentComposite?Conference=Big-Ten

This gives a great idea of talent actually on the roster, rather than just looking at the last 4 years of recruiting class rankings. This only includes the rating of the players on the team for that year. It does not include players that left the program, were kicked off, etc, that may have bumped up the rankings for previous recruiting classes.

For the Gophers I factored in all games including non-conference and bowl games:

2015
vs more talented teams: 2-7
vs less talented teams: 4-0

2016
vs more talented teams: 3-4
vs less talented teams: 6-0

2017
vs more talented teams: 2-4
vs less talented teams: 3-1

Total record vs more talented: 7-15
Total record vs less talented: 13-1

Overall, the Gophers have a winning percentage of 32% vs teams with more talent, and 93% against teams with less talent.

Then I looked at only the B1G head to head including the B1G championship game. I didn't do non-conference and bowl games due to time, but I don't believe it would change the outcome much.

2015
Record was 39-18 in favor of more talented teams

2016
Record was 47-17 in favor of more talented teams

2017
Record is 36-13 in favor of more talented teams

Overall the record is 122-48 in favor of more talented teams, for a winning percentage of 72%.

So on average the team with more talent won't win every time, but they'll win 3/4 of the time. This is why recruiting/talent matters.
 

How do you determine what a talented team is. We can all claim to be talented.
 


OK cool so coaching doesn’t matter. I knew we shoulda kept Brewster. Can’t believe Mase wasted time with guys like Barber, Decker, Eslinger, etc. Buncha 2 stars that did nothing. Also 2star Eric Murray blocked a field goal tonight
 

Since there are still several people on this site that don't believe in recruiting rankings, or that talent matters, I'll try to make it even more simple to understand.

You continue to oversimplify what people say about rankings and you know it. Then you accuse Swingman of having an agenda. Do you not realize that you are so transparent?


Sent from my iPhone using Tapatalk
 


Since there are still several people on this site that don't believe in recruiting rankings, or that talent matters, I'll try to make it even more simple to understand.

247 has a yearly team talent composite dating back to 2015. https://247sports.com/Season/2017-Football/CollegeTeamTalentComposite?Conference=Big-Ten

This gives a great idea of talent actually on the roster, rather than just looking at the last 4 years of recruiting class rankings. This only includes the rating of the players on the team for that year. It does not include players that left the program, were kicked off, etc, that may have bumped up the rankings for previous recruiting classes.

For the Gophers I factored in all games including non-conference and bowl games:

2015
vs more talented teams: 2-7
vs less talented teams: 4-0

2016
vs more talented teams: 3-4
vs less talented teams: 6-0

2017
vs more talented teams: 1-2
vs less talented teams: 2-1

Overall, the Gophers have a winning percentage of 28% vs teams with more talent, and 92% against teams with less talent.

Then I looked at only the B1G head to head including the B1G championship game. I didn't do non-conference and bowl games due to time, but I don't believe it would change the outcome much.

2015
Record was 39-18 in favor of more talented teams

2016
Record was 47-17 in favor of more talented teams

2017
Record is 18-4 in favor of more talented teams

Overall the record is 104-39 in favor of more talented teams, for a winning percentage of 73%.

So on average the team with more talent won't win every time, but they'll win 3/4 of the time. This is why recruiting/talent matters.

Great post. I would be very interested to see the same analysis done on a larger dataset - it's a shame this metric only goes back three years.
 

OK cool so coaching doesn’t matter. I knew we shoulda kept Brewster. Can’t believe Mase wasted time with guys like Barber, Decker, Eslinger, etc. Buncha 2 stars that did nothing. Also 2star Eric Murray blocked a field goal tonight

We won a lot of music city bowls with those guys.
 


We won a lot of music city bowls with those guys.

That’s what I’m saying. Who wants to win music city bowls stars when you can win 3 games and lose to south Dakota on your own field with higher ranked talent
 



That’s what I’m saying. Who wants to win music city bowls stars when you can win 3 games and lose to south Dakota on your own field with higher ranked talent

Those aren't the only two options.
 



For the Gophers I factored in all games including non-conference and bowl games:

2015
vs more talented teams: 2-7
vs less talented teams: 4-0

2016
vs more talented teams: 3-4
vs less talented teams: 6-0

While I don't think this metric exists for 2014, I took a look at the 2014 schedule in a similar manner:

Less talented:
Eastern Illinois (WIN)
Middle Tennessee (WIN)
San Jose State (WIN)
Northwestern (WIN)
Purdue (WIN)
Illinois (LOSS)

Record vs. less talented: 5-1

More talented:
TCU (LOSS)
Michigan (WIN)
Iowa (WIN)
Ohio State (LOSS)
Nebraska (WIN)
Wisconsin (LOSS)
Missouri (LOSS)

Record vs. more talented: 3-4

That means over a three-year period the Kill/Claeys regime was 15-1 (5-1 in 2014, 4-0 in 2015, 6-0 in 2016) against less talented teams and 10-17 (3-4 in 2014, 2-7 in 2015, 3-4 in 2016) against more talented teams. While the 15-1 record is obviously very impressive, I think the 10-17 record is impressive in its own way as well.

This speaks to the game-planning ability of the previous regime, IMO. Beat the teams you are supposed to beat, albeit sometimes in ugly fashion, and put up a fight against teams a step (or half-step) better than you.

That being said, as many have detailed, it is in fact possible to believe the previous coach was a good coach AND support the decision to fire him to make room for PJ. Some people still seem to be struggling with that concept.
 



Great post. I would be very interested to see the same analysis done on a larger dataset - it's a shame this metric only goes back three years.

I know, but I'm guessing it isn't easy to keep up with all the players coming and going. I think 247 will keep doing it every year on now.

Here's analysis from a few years ago. https://deadspin.com/chart-which-ncaa-football-teams-outplay-their-recruit-1640831522

Correlation between recruiting rankings and wins =.77, which again is strong.
 

Nobody is going to argue talent doesn't matter. It does. But coaching is important as well. The best programs tend to have the best coaches. Some coaches do better than expected, some do less than projected and it tends to continue that way until a coaching change (with exceptions).

Look at Michigan under Carr, then Richrod, then Hoke, then Harbaugh. Did they recruit poorly for 10 years? I doubt it. Institutional support lacking? Are you ****ting me? Right coach comes in and rights the boat but, he still needs a good QB, still needs a good line, etc.
 

For those of you pinning your hopes on "croot rankings" and lists, etc I've got some bad news...we're currently 9th in the Big Ten on average recruite composite. Now, I have no idea what the composite is, or where it comes from, and we've already shown it's some sort of half-baked significant figure-ignorant backwoods maths but I suppose if it's applied equally it might have some sort of validity as a measuring stick. Sadly, unless PJ show he can coach 'em up (and it looks promising so far with #8, for example) and out-strategize and out-tactic his opponents we are in for a long slog to the top (sound familiar?)

In 2017 we were 12th
2016 10th
2015 12th
2014 11th.
You can find the rest if you want.

Did we over or under achieve recruiting ranking composites under Kill, Mason, (Fleck)?
 

I don't think anyone is saying that recruiting doesn't matter, or that talent doesn't matter. Some of us just don't think that you can distill a successful program down to one answer.

A successful program is like a 500-piece jigsaw puzzle. you can't pull out one piece and say "this explains everything." There's recruiting, coaching, player development, offensive and defensive systems, putting the right players in the right places, conditioning, avoiding and/or preventing injuries, avoiding/limiting turnovers and penalties, facilities, administrative support, fan and booster support, and more.

It's great to have talent, but that talent needs to be coached. personally, I will take a good coach with lesser talent, as opposed to a poor coach with superior talent.

And, on the recruiting rankings - I maintain that is not an exact science. just because website X or recruiting service Y says this team is ranked 30th, and another team is ranked 40th - that does not mean the 1st team is guaranteed to be better. that is an on-going debate - how much real difference is there between the 30th and 40th best recruiting class, as opposed to the difference between the 10th and 20th recruiting class. Is it even possible to quantify? The 3* and 4* ratings are not magic numbers assigned by God. They are estimations of a player's talent and/or potential - assigned by humans, who may - or may not - be correct.
 

Nobody is going to argue talent doesn't matter. It does. But coaching is important as well. The best programs tend to have the best coaches. Some coaches do better than expected, some do less than projected and it tends to continue that way until a coaching change (with exceptions).

Look at Michigan under Carr, then Richrod, then Hoke, then Harbaugh. Did they recruit poorly for 10 years? I doubt it. Institutional support lacking? Are you ****ting me? Right coach comes in and rights the boat but, he still needs a good QB, still needs a good line, etc.

To expand on the bolded I think GWG (and I applaud your effort above but it's not drilling nearly deep enough) or someone else needs to plot the winning percentage vs last 5 recruit rankings (over any arbitrary timeframe):

1. First time head coaches
2. First time head coaches promoted within the same program
3. Established head coaches
Bonus:
4: G5 HCs
5. P5 HCs
6. Helmet vs Non-helmet program

Reason being, we know coaches like Urban, Saban, Patterson, Dantonio, Petersen, Harbaugh and some others have established track records as HC at a minimum two programs or a long and distinguished career at their current. It stands to reason a new or first time head coach would have the same success IF recruiting/talent is the only element in the formula.
 

If we can't row the boat to victory in four years, we will be forever platt, ehr fifth in the B1G.
 

You continue to oversimplify what people say about rankings and you know it. Then you accuse Swingman of having an agenda. Do you not realize that you are so transparent?

Since GWG put some effort into his analysis I thought I should do more with my response.

1. Why are we removing players that never made it to campus, are no longer with the team, etc.? That is a primary reason I don't put much stock in team rankings (see Brewster). When has that ever been discussed in the debates we have had here - including the most recent one that started again with the MD debacle? Removing that data is presented here (and by 247) as some way to be more accurate - but it does nothing but invalidate the data when the argument is about recruiting rankings.

2. Few, if any, in the group would debate that an Ohio State team ranked in the top 5 isn't better than a Rutgers Team ranked 75. That has never been the debate (for me at least). My point has always been after the top 20 or 25 or pick a number around there - it is a roll of the dice. The same happens at the back end too, of course. The ranking police would basically argue the team ranked 32nd is better than that ranked 33rd and my argument is I wouldn't even assume that the 25th ranked team is better than the 50th. I wonder how this data would look if we only used teams between 20-60 or 25-75, or something like that? Including the extreme ends is, again, including data not part of the central argument to influence the result.

Sent from my iPhone using Tapatalk
 

Another angle to look at with recruiting rankings is helping each individual class. Lets say Quinn Carroll commits to the Gophers. He is a high 4 star guy OT. If he commits, it would only benefit the recruitment of the 4 star QB out of Southern California. Lets say he commits, that could entice a higher rated WR to commit, and so on. These kids see the stars of other kids in a school's recruiting class, and I bet it has some weight on their decision to commit.

A different topic to think about is the advantage of having kids commit and start graduate early and are early enrollees. Not only are they getting into the program that much earlier and can participate in spring ball, but obviously good enough students where academics are less likely to be an issue.
 

I don't know about everyone else, but I took these stats to mean that if you want to be consistently good and compete, you need more talent. Good coaching is valued, but that's just one piece. We need the stronger recruiting classes to have those breakthrough seasons. Based on the statistics, it's reasonable that every now and then, a "less-talented" team could have an incredible season (thanks to good coaching, luck, etc). But it's highly unlikely.

I hope Fleck has the ability to bring in the required talent.
 

There are a whole lot of variables to the overall equation; however, GWG has provided quality data that supports the overall assertion that talent, indeed, matters not just a little, but a lot. (Thanks for the work on this, GWG.)

...Not that that is some big surprise...

Fleck, so far, has been unable to bring in the A+ talent but he appears to be bringing in depth of talent the likes of which we haven't seen much of over the past few decades, particularly those kids with multiple high-quality offers. Instead of competing with Central Michigan and NDSU and Middle Tennessee State for players, we're now competing with bigger schools; given that, I'm surprised that the players who have committed thus far have all been limited to "3*s".
 

Since GWG put some effort into his analysis I thought I should do more with my response.

1. Why are we removing players that never made it to campus, are no longer with the team, etc.? That is a primary reason I don't put much stock in team rankings (see Brewster). When has that ever been discussed in the debates we have had here - including the most recent one that started again with the MD debacle? Removing that data is presented here (and by 247) as some way to be more accurate - but it does nothing but invalidate the data when the argument is about recruiting rankings.

2. Few, if any, in the group would debate that an Ohio State team ranked in the top 5 isn't better than a Rutgers Team ranked 75. That has never been the debate (for me at least). My point has always been after the top 20 or 25 or pick a number around there - it is a roll of the dice. The same happens at the back end too, of course. The ranking police would basically argue the team ranked 32nd is better than that ranked 33rd and my argument is I wouldn't even assume that the 25th ranked team is better than the 50th. I wonder how this data would look if we only used teams between 20-60 or 25-75, or something like that? Including the extreme ends is, again, including data not part of the central argument to influence the result.

Sent from my iPhone using Tapatalk

That's an excellent point Spoof, I wonder if there is any data to show that?
I'm with many of you here, while I would love to see a top 25 class what matters to me the absolute most is the number of offers from other P5 schools. That tells me that the kid has talent and if he is not a 4* (or 5* for that matter for those ranking using that matrix) that he is a very worthy recruit. I love reading about the crootin' though from those in the know, I find it very interesting. So far it appears PJF is finding most of his croots to have a lot of P5 offers. I like that!
 

Since GWG put some effort into his analysis I thought I should do more with my response.

1. Why are we removing players that never made it to campus, are no longer with the team, etc.? That is a primary reason I don't put much stock in team rankings (see Brewster). When has that ever been discussed in the debates we have had here - including the most recent one that started again with the MD debacle? Removing that data is presented here (and by 247) as some way to be more accurate - but it does nothing but invalidate the data when the argument is about recruiting rankings.

2. Few, if any, in the group would debate that an Ohio State team ranked in the top 5 isn't better than a Rutgers Team ranked 75. That has never been the debate (for me at least). My point has always been after the top 20 or 25 or pick a number around there - it is a roll of the dice. The same happens at the back end too, of course. The ranking police would basically argue the team ranked 32nd is better than that ranked 33rd and my argument is I wouldn't even assume that the 25th ranked team is better than the 50th. I wonder how this data would look if we only used teams between 20-60 or 25-75, or something like that? Including the extreme ends is, again, including data not part of the central argument to influence the result.

1. There's going to be losses at every school from every recruiting class. Some are hit harder than others, but even after you exclude those losses, they're still going to fall in line with where their recruiting classes averaged. The Gophers would have probably been ranked closer to 10th this year, but with their defections from the 2014/2015 classes, that dropped them to 12th. Still, even the hardest hit teams are unlikely to move more than a spot or two if their classes are consistently in the same range. Michigan St is another example after losing several highly rated players. It dropped them from 4th to 6th.

2. So lets take the Gophers record vs the teams in that 25-50 range on a yearly basis. That's most of the teams in the west. Wisconsin, Nebraska, Iowa, Northwestern. Other teams in that range, Rutgers, Maryland, and Indiana, but we've only played them once or twice the last 6 years.

Head to head the last 6 years: Wisconsin 0-6, Nebraska 2-4, Iowa 2-4, Northwestern 3-3. Total: 7-17. Winning percentage: 29%
The correlation is still very similar to my original post.

If you want to look at just the last 2 years since that was the data I pulled, it's: Wisconsin 0-2, Nebraska 0-2, Iowa 0-2, Northwestern 1-1. Even being consistently 10 spots higher on a yearly basis matters.
 

Give me great S&C and player development. I'll take a great winning culture, top coaching, great team chemistry and detailed execution to regularly beat your 3 stars with 2 stars. Rankings don't mean squat without these key pieces in place in the ultimate team sport, especially in the trenches. Rankings do not take style of play into account either.

If you think it's all about rankings, you are sadly mistaken.
 

A case study in how to bend statistics to fit a narrative. In general, talent helps but there are outliers above and below expected based on other factors. If one invests too much one way or the other it's easy to overlook red flags.

If one searches the nets long enough someone did a retrospective look at individual programs actual wins vs expected and it was an enlightening piece.
 

A case study in how to bend statistics to fit a narrative. In general, talent helps but there are outliers above and below expected based on other factors. If one invests too much one way or the other it's easy to overlook red flags.

If one searches the nets long enough someone did a retrospective look at individual programs actual wins vs expected and it was an enlightening piece.

No one’s bending statistics. This analysis has been done over and over again by many different sources. Talent above all else is the #1 predictor of success. Continuing to argue against simple statistics is wrong.
 

No one’s bending statistics. This analysis has been done over and over again by many different sources. Talent above all else is the #1 predictor of success. Continuing to argue against simple statistics is wrong.

I'm not arguing against it.
 

No one’s bending statistics. This analysis has been done over and over again by many different sources. Talent above all else is the #1 predictor of success. Continuing to argue against simple statistics is wrong.

Jimmys and Joes over Xs and Os. Who is saying talent doesn't matter? Isn't the debate about using recruiting rankings as the gospel in regards to talent? And, btw, I already pointed out a couple ways you did "bend" data to fit your agenda - but you chose to dismiss them. That's fine, but not everyone has to agree with you. That's fine too.


Sent from my iPhone using Tapatalk
 




Top Bottom