The following is a comprehensive guest post from internet friend, Jason. He’s a regular contributor at All Buffs and my gchats.

Deep benches in basketball seem to be a recipe for success.  The term “depth” gets thrown around like a double play. It gives you line-up options, protection against foul trouble or injury; the ability to go all out and press or play pressure D for a full 40.  This all seems to make sense, it’s intuitive, but does it actually correlate to winning? Because winning is the best stat.

There are a couple ways to look at this. Does a long bench help you through the season more than in the NCAA tournament? Or is the reverse actually true? Looking at the final 4 teams for the last 5 years, there have only been 2 teams (2013 Wichita St and 2009 Michigan St.) in the top 100 in the percent of bench minutes played (thanks KenPom). And just two Final Four teams ranked in the 100’s (2013 Louisville and 2010 Michigan St).  Meanwhile, there have been nine in the 300’s. It appears that having a short bench full of go-to guys means more than having a bunch of interchangeable guys in the NCAA tournament.

NCAA Final 4 Tournament teams overall ranking in % of bench minutes played

The reality of this isn’t terribly shocking. The best players seem to win the most games. That 2012 Kentucky team that has set the bar for modern, single season domination had six players drafted. Coach Cal called upon his reserves the 323rd most frequently. They beat you because they were good, not deep. In a three week sprint for the finish, you don’t necessarily need more, just the best.

But what about the other side? Garnering an invitation into the Dance is quite an accomplishment. Tourney appearances hang from rafters, too, so it’s worth exploring whether or not, across the grind of thirty-plus games including a conference slate, a bench is helpful. The inevitability of injuries, sickness, wear-and-tear, the dog days, and every other unpredictable occurrence. Does having a deep bench help spell those issues?

Again, the answer appears to be “no”.  Over the past 3 seasons in the Pac-12 there is a negative (-0.213) correlation between bench minutes and winning. This year the overall NCAA is on pace to have an overall negative correlation higher than the three year Pac-12 average of -0.236.

Some of the deepest teams have actually been the worst. Take this year’s Oregon St. team: they have the deepest bench in the three year study of the Pac-12. Craig Robinson’s playing his bench 39.2% of the time and it’s  earned him 10^th place in the standings.  The conference regular season winners are 11^th, 10^th and 10^th in the percent of bench minutes played in the three years of Pac-12 hoops studied. Conversely, the longest benches have finished 10^th, 3^rd and 7^th.

Bench Min by Year, 3 year avg, and avg Pac-12 regular season conference finish

Scouring over this we may be inclined to do a case study on Dana Altman. His teams have the highest average % of bench minutes played in the 3 year history of the Pac-12. There are no outlier years. He consistently plays his bench a lot.  This makes some sense on the periphery; he brings in a lot of transfers, trying to figure out lineups that work, tossing out a lot of different combinations and bench players.  But if you dig a little deeper you can see that even at Creighton he had a run of three consecutive years where he had the 3^rd highest percent of bench minutes played. Clearly he’s a guy that likes to go deep into his bench and he’s only able to partially execute that plan with the roster turnover he’s had at Oregon.

Of note, and considering our first point with regards to Final Fours and our second note with regards to winning conference titles, Altman has been to just one Sweet Sixteen and only won or shared three conference titles in twenty-five years as a head coach.

In theory, having a long bench is terrific. You can spell all those season long unexpecteds. But, like I science experiment, perhaps introducing too many variables gets you into a situation where you can’t control the results. A classic case of less-is-more, maybe the bench is just for the Bernie: