UConn True Shooting Through Six Games

After looking at some of the shot location trends as the game goes along for the UConn basketball team, I decided to look at each players shooting efficiency through six games. Shooting efficiency can be summed up nicely in one number, true shooting percentage. True shooting percentage is the ratio of points scored to true shooting attempts, where a single free throw is 0.44 true shooting attempts (this figure was derived from considering all free throws, i.e. those yielded after two point fouls, three point fouls, and free throws after a made basket). All field goal attempts represent one true shooting attempt. The benefit of looking at shooting through the lens of true shooting percentage is it adjusts for the fact that a three pointer is more valuable then a normal two point field goal and it rewards players who get to the free throw line.

The following table represents the shot profiles of all the relevant players in the UConn rotation:

Vital has put up the most true shooting attempts through six games followed by Carlton and Gilbert. Given Gaffney's lack of playing time and lack of aggressiveness with regards to looking for his shot, it is no surprise that he has shot less than half as many true shooting attempts as the next closest player in the rotation. Below is a plot that compares efficiency, in the form of true shooting percentage, and volume, in the form of true shooting attempts:

The horizontal and vertical lines represent the averages in efficiency and volume respectively. You can see that with his high volume, Gilbert has not been efficient shooting the ball thus far. Now, this will regress some, a 36.5% true shooting is not sustainable for a player of his capability. This also neglects to account for the shots he has created after drives to the rim, where he has the ability to blow by defenders with his elite quickness. Polley has maintained very good efficiency with his proficiency from three point range. James Bouknight has only played in three games (hence the low volume) but has shot the lights out thus far. The same can be said for Brendan Adams after a poor freshman year. Vital has maintained solid efficiency despite the degree of difficulty of his shooting attempts. Carlton's similar efficiency to Vital is disappointing when you take into account the proximity to the basket with which he takes his shots and the fact that many of his attempts are assisted. Sid Wilson's lack of offensive production has to be considered a let down given his recruiting pedigree. If he can at least become a passable shooter with his athleticism, he could be a really important piece for this UConn team. Finally, here is the same plot but for just the last 10 minutes of games:

This information should be taken with an extremely sizable grain of salt, given the lack of game time in this split. Bouknight and Adams have shot extremely well at the end of games. Given they are both perimeter based players, this will regress some. Vital has taken his game to another level during this time. Carlton and Gilbert lead the team in volume at the end of games, yet they have been two of the least efficient players in this spot. I would imagine the two of them will have a little more luck finding the bottom of the basket as the season goes on.

Thank you to Luke Benz and his ncaahoopR package for making it easy to scrape the play-by-play data

UConn Shot Profile Thus Far

I was looking at the play-by-play data of all the UConn games through today (11/24 against Miami Fl.) and I noticed what I thought was an odd trend:

I filtered out garbage time (win probability between 15 and 85 percent) when looking at this data. Anecdotally I would think the trend would be the reverse; as the game goes on the defense makes more of an effort to wall off the paint. However, as the game has gone on for the Huskies thus far, they have dramatically decreased the rate of three pointers taken. This can be attributed somewhat to what seems like more post-ups for Josh Carlton in close games (to my chagrin) and taking the ball out of the guards hands more than what would be ideal. Here are who has taken the shots in the last five minutes of the game:

Josh Carlton is just third on the team in shots taken in the last five minutes and of course has taken no three point shots. All of Tyler Polley's seven shots have been threes. Interestingly enough, Vital has close to a 50% three point attempt rate overall but in the last five minutes only has a 33.3% rate, a very abrupt decline. Bouknight has only played in three games so far, but based on the way his shot looks, I hope in the future he becomes more aggressive from threes in tight games (note that his entire sample of shots in the last five minutes has been in close games).

Now, it's worth noting, this extremely small sample of last five minute shots does not give this much predictive value. But the team's shot profile overall as a function of time thus far I thought was worth noting. It will be interesting to take a look at this in a few months.

Thanks to Luke Benz for making it easy to scrape the play-by-play data

Gerrit Cole's Free Agency

As any baseball fan with a pulse knows, Gerrit Cole is set to cash in this off-season. He will probably sign off on the largest free agent contract for a pitcher in league history. While we can all wait and guess where he is going to go and for how much, I figured the conclusion of the World Series would be a great time to project how much he will be worth on the mound over the life of his contract.

When assessing Cole's value (as a pitcher), I assumed that he would sign an eight year contract. Next season will be his age 29 season, so an eight year deal that takes him through his age 36 season seemed reasonable given his pitching prowess and the state of the free agent market (news flash: teams are becoming increasingly weary of handing out long-term deals through a player's 30's).

In order to project his value over the life of the contract, we need to estimate Cole's true talent as a pitcher right now and then apply some sort of technique to reflect aging season over season. First, to get an approximation of  Cole's true talent right now, I took a weighted average using the weights from Tom Tango's Marcel projection model (five for this past season, four for the season before, and three for the season before that). Using FanGraphs WAR, Cole's "true talent" right now is a shade under six wins (5.93 to be exact).

Once I obtained this value, I applied a projected change in WAR for each of the eight seasons. While somewhat arbitrary, I considered changes in performance in line with the results of more rigorous  studies into aging players. I considered a -.25 WAR drop off in his age 29 season, -.5 WAR for ages 30, 31, 32, 33, and 34, and -.75 for ages 35 and 36. However, I also wanted to get a range of outcomes within each projected year-to-year decline. So for each season, I created a random normally distributed set of numbers with a mean of the deltas I described above and a standard deviation of one. Each year I pulled out the fifth, twenty-fifth, fiftieth, seventy-fifth, and ninety-fifth percentile deltas and added those to Cole's true talent figure. This would provide a distribution of outcomes for Cole's value to whatever team signs him. I then summed up the WAR values in each case and multiplied them by 9, 9.5, and 10 which represents estimates for the cost of a win on the free agent market. Here are the results:

For more specificity, here are the actual figures:

In close to the reasonable best-case scenario, Cole is projected to be worth a bit over 400 million dollars by my methodology. Correspondingly, in about the worst-case scenario, he figures to be worth about 170 million. Now, it is most important to look at the fiftieth percentile outcome, which is supposed to represent what a team should be realistically expecting from Cole. His value in this case would be about 300 million over the course of an eight year contract.

Now, given the state of the free agent market, I would be surprised if Cole signed for 300 million dollars. So what can we conclude about whatever team signs Cole? That team is probably getting an excellent pitcher on a bit of a discount.