Syracuse Basketball: The Good and Bad of Advanced Metrics

Nov 17, 2015; Syracuse, NY, USA; Syracuse Orange guard Franklin Howard (1) passes the ball around St. Bonaventure Bonnies forward Derrick Woods (20) during the first half at the Carrier Dome. Mandatory Credit: Rich Barnes-USA TODAY Sports
Nov 17, 2015; Syracuse, NY, USA; Syracuse Orange guard Franklin Howard (1) passes the ball around St. Bonaventure Bonnies forward Derrick Woods (20) during the first half at the Carrier Dome. Mandatory Credit: Rich Barnes-USA TODAY Sports /
facebooktwitterreddit

Advanced metrics are a great way to tell how efficient a player is. For Syracuse basketball, they can tell us a lot about the team, but they can also be incredibly misleading.

As usual, I like to browse around random Syracuse basketball message boards just to see what the buzz is. And as usual, it’s a jumbled mess of opinions and trolls. “Cooney is awful!” “Boeheim can’t coach, he needs to go!” “Why isn’t Obokoh playing!?!”

More from Syracuse Basketball

On and on they go. I find it quite amusing to watch the arguments unfold between superfans who all think they know more than everyone else and that their opinion must be right. I get it. Heck, I sometimes feel that way too. I try to be a bit less aggressive about it and usually like to at least listen to another argument, but I get it. Everyone wants to think they are infallible in their sports knowledge.

One of the things that bugs me the most however, is when people point to advanced metrics and use that information as if it is fool proof. Like, because you used a certain advanced metric, it means that no other statistic is valid.

First off, let me just say this. I love advanced metrics for the most part. They are an excellent way to tell exactly how a player is performing on a per minute basis, they can show how efficient a player is on offense and defense, and they allow a much greater statistical breakdown of what a player does well and where he struggles.

In case you’re unfamilar with the term, advanced metrics is basically the use of situational data in a game to come up with some additional barometers to measure how good a player is outside of just how many points they score or rebounds they grab.

It can be as simple as finding out the average number of points/assists/rebounds/etc per 40 minutes (for college, 36 is the standard for NBA), and as detailed as determining how effective a defender is based on how many average points he gives up at his position per opponent’s possession, how many defensive stops he’s responsible for, or the team’s offensive efficiency when he’s on/off the court.

It can be overly complicated at times. If you want a better breakdown of what it means, here’s a great resource to give you some more of the details.

Anyways, getting back on track. I happened to be browsing on syracuse.com, one of my favorite spots to sit down with a bag of popcorn and just laugh at the rampant flame wars between posters. One such war really grabbed my attention, enough to the point that I wanted to write this article.

Basically, a fan was saying that based on advanced metrics, Frank Howard is a better player than Trevor Cooney. The metrics they used to back this up were based on per-40 minute stats such as assists per game, rebounds per game, steals per game, turnovers per game, and assist/turnover ratio.

Syracuse Basketball Franklin Howard
Dec 8, 2015; Syracuse, NY, USA; Syracuse Orange guard Franklin Howard (1) brings the ball up court during the first half of a game against the Colgate Raiders at the Carrier Dome. Syracuse won 78-51. Mandatory Credit: Mark Konezny-USA TODAY Sports /

For comparison sake, here are those numbers, just so you’re aware.

Trevor Cooney, per 40 minutes, averages 14.2 points, 2.6 rebounds, 2.7 assists, 1.7 steals, and a 1.47 assist/turnover ratio.

Frank Howard, per 40 minutes, averages 6.3 points, 5.8 rebounds, 6.9 assists, 1.4 steals, and a 2.56 assist/turnover ratio.

More from Inside the Loud House

Looking at those stats, you’d think, wow, why doesn’t Howard play way more and Cooney play way less? I mean, while the point differential favors Cooney, almost every other stat favors Howard. Clearly that means Howard is the more well-rounded player, right?

Not necessarily.

First and foremost, just focusing on these specific advanced metrics isn’t the end-all be-all. It doesn’t take into account defense (other than steals), which Cooney wins by a large margin. It also doesn’t even add in shooting percentages, where Cooney is much better from 3 and the free throw line. It doesn’t account for many of the intangibles that don’t fit specifically into a per game statistic such a defensive stops, defensive field goal percentage, opponents points per possession while the player is on the floor, team offensive efficiency while on the floor, etc etc.

The even bigger issue though, is that Frank Howard only plays 10 minutes per game. Extrapolating that out to 40 minutes is making a lot of assumptions. I mean, just look at those assist and rebound statistics.

According to those numbers, Howard would be #1 in the ACC in assists per game… by a lot. James Robinson leads the ACC at 5.3 assists. Even if Howard played just 30 minutes per game, based on those numbers, he’d still be second in the ACC in assists per game. He’s good, but he’s not that good.

Rebounding is just as far-fetched. Based off those per-40 numbers, Howard would average more rebounds per game than Michael Gbinije…. again, by a lot. Howard is a decent rebounder, but he’s no Gbinije. He’s 4″ shorter and not in the same league with him in that category.

In fact, based on those per-40 numbers, Howard would be Syracuse basketball’s top rebounding guard in the past 30 years on a per game basis. Yes, you read that right. Going beyond just guards, he’d be averaging more rebounds per game than players like MCW, CJ Fair, Rakeem Christmas, Otis Hill, and others did in their careers.

Next: SU Basketball vs. NC State: Game Preview

That, my friends, is why you can’t just go by advanced metrics as the only criteria. Or if you do, you need to factor in things aside from just per-game statistics, and also factor in minutes played. Assuming that a guy who plays 10 minutes a game will be exactly 4 times as productive in 40 minutes is asinine. Maybe Howard is a much better scorer in more playing time. Maybe he turns it over way more. Perhaps he’s a much better shooter. We don’t know, and making those assumptions off of such a small sample size is just not accurate.

Whew, well that got a bit longer and a bit more in-depth than I planned. Anyways, the whole point of this article is to say that yes, advanced metrics can be an excellent way to learn how efficient a player is. But using them as the sole gauge of player performance and potential is inaccurate and leads to long-winded articles from yours truly.

Don’t be that guy.