On IPCs and Football Scholarships

Submitted by EGD on January 29th, 2013 at 3:32 AM


As any chess master will tell you, allowing one’s pieces to languish on the back rank is a certain invitation to humiliating defeat.   Or, as anyone who has ever played Axis & Allies well knows, a general who hordes his Industrial Production Certificates will quickly fall to the opponent who transforms hers into military units and strategically deploys them as rapidly as possible.  So I was quite surprised the other day when several posters suggested that Michigan ought to “bank” its last remaining scholarship for the 2013 class, rather than sign a player who might not arrive in Ann Arbor with the highest of expectations.  Like chessmen or IPCs, I have always felt that a football coach must aggressively leverage production from his full complement of 85 scholarships—or as many of that number as possible—if he hopes to outcompete the other 120 programs in the country. 

Scholarships are not chessmen, of course, nor are they IPCs—and the calculus that goes into offering and signing a collegiate student-athlete is quite a bit different than the evaluation of choices in a board game.  So, although the idea of purposefully letting a scholarship go unfilled struck me as intuitively unwise, the suggestion did not seem entirely without merit.  After all—if by passing on a probable depth player in 2013, Michigan could sign a likely frontline contributor in 2014, then the payoff might be worth the investment.  The overarching strategic principles remain sound and generally applicable, I felt, but is this case the exception?  I decided to take a closer look. 

Most estimates place the expected size of Michigan’s 2014 class at around 17 scholarships.  This projection appears based on fifteen players exhausting their eligibility in 2013, and two redshirt juniors not being offered fifth years.  With usual attrition, UM would more realistically expect to have about 20 scholarships available.  But for purposes of this analysis, I will presume the 17 figure holds true.

Banking a 2013 scholarship would enable UM to sign an 18th player in the 2014 class.  Therefore one part of the “to-bank-or-not-to-bank” equation seems to be the reasonably anticipated quality of the eighteenth recruit in UM’s 2014 class.

In 2012, Michigan had twelve recruits who received four or five stars on Rivals; the 18th-highest recruit would have been a 3-star with a 5.7 grade (Ben Braden, Jeremy Clark, Devin Funchess, Matt Godin, Mario Ojemudia, Kaleb Ringer, AJ Williams, and Chris Wormley fit that description, according to Rivals).  In 2013, Michigan has seventeen recruits with at least four stars and a 5.8 grade, though the 18th-best recruit again checks in with 3 stars and a 5.7 grade.  Therefore, Michigan’s performance in the last two recruiting cycles would seem to suggest that banking a 2013 scholarship would most likely produce a high (5.7) 3-star recruit to Rivals. 

A slightly improved performance in 2014 could realistically land a low (5.8) 4-star recruit.  However, between 2005 and 2012 only twelve recruiting classes have featured at least 18 players rated four stars or higher on Rivals.  And, of those twelve classes, all but two (2006 FSU and 2008 ND) belonged to teams that had won (mythical) national titles within the preceding decade.  Michigan, which hasn’t won the MNC since 1997 and hasn’t seriously contended for one since 2006, and which doesn’t happen to be in a talent-abundant state like Florida or have a Catholic pipeline like ND, would not seem especially well-positioned to defy this pattern.  On the other hand, Scout.com (which is a bit more generous with their star rankings) lists fifteen classes with eighteen or more 4+stars from 2005-12, and is already projecting two more for 2013: one of which belongs to Michigan.

We can thus assume that a hypothetical 18th recruit for 2014 would likely be a player on the 3-4 star borderline, with a Rivals grade of 5.7 or 5.8.  Maybe not a heavy bomber, but still a pretty high-quality recruit either way.  The drawback, of course, is that player would not join the team until the 2014 season. 

While landing eighteen or more 4-stars is uncommon—and almost unheard-of for non-MNC contenders—landing twenty-seven or more 3-stars ain’t no thang.  Between 2005 and 2012, there were 32 recruiting classes of 27 or more players that finished with a top-15 ranking on Rivals.  Of those classes, 18 (or 56.25%) had at least 27 players rated 3-stars or higher.  Now, if we were at the beginning of the 2013 recruiting cycle and were trying to predict Michigan’s chances of signing at least 27 three-stars, this percentage would already suggest Michigan has pretty good odds of pulling it off.  But with UM having already obtained verbals from 27 players, of whom 26 are rated 3+ stars (the 27th recruit is LS Scott Sypniewski)—and needing only to fill the one remaining spot, the chances of Michigan being able to find one more 3-star recruit for that final spot would appear to approach 100%. 

So, let’s assume for purposes of the remaining discussion that the final spot would to go a (mid) three-star player with a Rivals grade of 5.6.  This player would be a tad less talented than the hypothetical 2014 signee, but would have one more year of experience in the program.  Assuming one year of collegiate coaching and strength & conditioning is equal to or greater than the value of a .15 upgrade on the Rivals grading scale, recruiting a slightly less-talented player in 2013 is at worst equivalent to signing a more talented player in 2014 (as the 2013 player’s redshirt season cancels out the banked-scholarship season for a recruit who plays as a true freshman).  But, assuming both players would redshirt their first years in the program, UM would sacrifice an entire season of production from one scholarship position.  It is doubtful that the slightly greater overall production one might expect from a 2014 player over the course of his career would sufficiently off-set this high immediate cost.


Player #28 of 2013 Class

Player #18 of 2014 Class


Redshirt (no production)  or 5.6

X – No Production


5.6 + 1 year

Redshirt (no production) or 5.7


5.6 + 2 years

5.75 + 1 year


5.6 + 3 years

5.75 + 2 years


5.6 + 4 years or new recruit

5.75 + 3 years


New recruit

5.75 + 4 years or new recruit


In the end, while there appears to be a stronger case for banking the last scholarship than I expected, I still think UM ought to take another player if they can find a Willie Henry or Dennis Norfleet type of player to come on board.  The potential benefits of saving the scholarship for 2014 are tenuous and distant, while the costs are immediate and certain.  Moreover, any attrition that occurs between now and 2014 will further diminish the expected returns from the hypothetical 2014 player, as a 19th recruit taken in 2014 is presumably less likely to be of four-star quality than the 18th, a 20th player even less, and so on.  Then, of course, some of the most likely targets for that final spot are defensive tackles, who would become subject to the Heininger Certainty Principle (which, frankly, is better than super submarines, long-range aircraft, or even V-2 rockets!).  

[Edit: Nerd that I am, I suppose I subconsciously view Axis & Allies as essentially a gen-X Monopoly or Clue—that is, a board game title with which anyone within, say, ten years of my (37 year-old) vintage ought to be reasonably familiar.  Thus, as I was trying to finish this diary in the wee hours last night, I evidently didn’t think it necessary to include a brief primer on what A&A actually is, or the basic strategy underlying the game-play.  Having thought better of the matter this morning, I offer the following supplement.

A&A is a famously-imbalanced, turn-based World War II strategy game that involves five players on two de facto teams: the Axis (with one player controlling Germany and the other Japan) and the Allies (comprised of players for the USA, UK, and Russia).  Each turn follows a pre-set sequence: Russia plays first, followed by Germany, then UK, then Japan, then USA.  Within each turn, a player first “purchases” military units using the currency of the game, Industrial Production Certificates (or “IPCs”).  A player receives IPCs by controlling territories on the game board (generally speaking, the more heavily industrialized the territory is, the more IPCs it is worth).  Once the units are purchased, however, they may not be deployed until the end of the player’s turn.  In the meantime, the player may maneuver his units and attack opponents—but only using his or her pre-existing units. 

Industrial Production Certificates are collected at the very end of a player’s turn, after all movement and combat has taken place.  And the amount of IPCs a player has to spend at the beginning of his or her next turn may be reduced through bombing raids or rocket attacks that other players launch in the meantime.  Furthermore, new units may only be deployed in spaces where a player has an “industrial complex”—often the only such space is a player’s home country—so it may take one or more future turns for a newly-deployed unit to travel to a forward area of the board where it can make a meaningful contribution to the game.  Therefore, it behooves any player to spend his or her entire allotment of IPCs at every opportunity—thus converting them into actual units that can occupy, defend, or invade territories (thus preserving or increasing a player’s future IPC stream).]



January 29th, 2013 at 8:33 AM ^

Thanks for your careful reasoning. I agree that one must calculate in the "time-value" of the recruit, and a dollar in the pocket today is worth more than the promise of a buck tomorrow. This time-value is lessened, however, by the need to somewhat "balance" classes so that, as we've had recently, you don't end up going back and forth with a huge class followed by a moderate class. Chess pieces don't run out of eligibility and fall off the board to be replaced. If one is a mid-level school hoping for a break-out season once every four years, bringing in a huge class may make sense. For an established program that wants to "re-load" (damn, I hate that expression, sorry), balance between classses has value.

That said, I think one wrinkle is that the player needs and players available for 2014 are only moderately known and knowable. The needs for 2013 and available players are much more certain. If Hoke feels that he still has a need for 2013 and there's a good fit, it would make sense to pull the trigger. On the other hand, even if your analysis is correct, if we cannot find a solid high-3 star or higher at a position of need for 2013, then banking the scholarship becomes the obvious choice.


January 29th, 2013 at 12:08 PM ^

I appreciate your comment, ChiBoyBlue.  But I think you are comparing the wrong things.  It is true that chess pieces don't exhause eligibility and fall off the board, but neither do scholarships.  Players exhause eligibility, but I think the correct view is to treat the scholarship as the asset and the player as the value produced by that asset.  You want as many of your 85 assets (scholarships) producing as much value (player-production) for as much time as possible.

With that in mind, I can see the attraction of having balance bewteen rrecruiting classes so that you can maintain a steady level of team quality from one season to the next.  But I doubt that the benefit a team would realizes from that kind balance could even be worth holding back scholarships for the future--assuming, of course, that a team doesn't have to fill the last remaining spots with athletes who are true long-shots to contribute.


January 29th, 2013 at 12:49 PM ^

You're correct that scholarships don't fall of the board, but the assets produced do. And there are a limited number of slots for those assets to fill, so there is a mandated redundancy. In other words, I'm not sure that chess is the best metaphor. It's more likely like a probability field, and the goal is to maximize the probabilities of success. In this case, a 5* has a better likelihood of contributing than a 3*, but sometimes he doesn't actually contribute. In addition, some positions can be filled successfully by a Frosh. Others (e.g., OL) generally require at least a redshirt year. Analyses of players tend to be more accurate for some positions than others. Some positions can be filled for an entire game or season by a single player, others require greater depth and substitutions. In short, I think the decision is based on far too many variables to cleanly analyze in this short a diary entry. I hate to say it, but this may be one where it starts heading toward apparent randomness, and we may be best trusting Hoke's (much revered) gut.


January 29th, 2013 at 1:17 PM ^

Yes, well, I certainly think you have a good point there.  But if we truly are confronting randomness, I see that as all the more reason to take a player this year rather than wait. 

The thing about chess is, high-level players like to assign relative point values to different pieces.  It varies depending on whose opinion you are looking at, but generally a pawn is considered to be worth 1 point, a minor piece (bishop or knight) about 3, a rook 5, and a queen 9 (the king cannot be exchanged and thus has an effectively infinite value).  Yet even these values are dynamic, in the sense that the placement of a particular piece and the situation/phase-of-game can make a piece mroe or less valuable.  For instance, a bishop that is trapped behind  one's pawns during a crowded middle-game might have an effective value below 3, whereas the same bishop on a long open diagonal might have an effective value above 3.  A rook has less practial value during the opening than the middle game, but becomes an extremely valuable piece during the endgame.  This, I think, makes the chess metaphor work in comparison with scholarships--which, like chess pieces, produce value at a fluctuating rate--but only so long as they are filled (i.e., "on-the-board").

Or, this whole thing could just be ridiculous.


January 29th, 2013 at 11:06 PM ^

I think this is a fascinating and well done thought exercise. You've broken it down as well as you can with the best knowledge (recruiting rankings) that the avergae fan has access too.  If I'm the coach, it boils down to if there are any guys left I think can be starters on Big Ten championship teams.  Not that the necessarily are going to reach that level, but that should be the aim. If I feel like a player's floor is a solid special teamer and his ceiling is that starter level, then you pull the trigger and risk not getting guy 18, 19, or 20. If not, then you just reward a deserving walk-on with a scholarship.


January 30th, 2013 at 12:53 AM ^

unless the coaches have the power to specifically control who commits when, then player 18 isn't necessarily of the caliber you describe. For example, if we have 17 guys signed up for next year, but some committed on later offers that went out when certain higher level prospects seemed to be inclined to go elsewhere, the risk might be not having the spot open for a last-minute high level player. So, say it's later in the process, and a 5.7 player at a position that takes time to develop has taken spot 17, but suddenly a 6.0 player that could start right away has late interest in us, and we don't have a spot for them, and thus turn them away. We offered the 5.7, who had just as real an offer as the 6.0 player, and they happened to take the offer earlier, and we cannot suddenly pull the offer because we think we can do better. Maybe this isn't the median scenario, but it seems plausible enough that I am actually forced to think randomness rules here, and it's probably just a coaching decision of who's available now versus who could be available next year.