Background: I profitably trade in the sports betting markets and for some reason decided to make a Twitter last May to share thoughts and ideas with one of the most biased, headstrong, negative and degenerate communities on the internet: #GamblingTwitter

Chief among the reasons I’m still here and enjoy it, are discussions with other analytical people who are smarter than I am.

If you follow me and are not already following Joe Peta, you are making a massive mistake.

I am currently in the midst of posting a thread about analytically interesting position players on each MLB roster and dedicated two days to discussing the defense and offense of one of my favorite athletes – Andrelton Simmons.

Peta replied to these threads with a link to his 2014 article about why defensive metrics overrate Simmons.

I was very surprised to see someone I respect so much disagree so strongly with me.

Here’s my response.

Don’t Come After My Man

The summary of the evidence Peta uses to call my hero a “homeless man’s Ozzie Smith” is this:

While Simmons converted more batted balls to outs than any other shortstop, it was not enough to be worth saving nearly 100 (5.5 dWAR * 9.264 RPW * 2.078 H/R = 105.8 hits) more hits than an average shortstop. (His exact numbers are a bit off, but the math above is using 2013’s league environment for Runs Per Win (RPW) and hits per run (H/R)).

My math does not match his – and if it did, I still think there are reasons to believe Simmons is a very wealthy man’s Ozzie Smith.

Using baseball reference’s Play Index, I found 2013 to have 15,305 total outs hit to shortstops – 12,068 groundouts, 2,311 pop flys, 924 line drives, and 2 bunt pop ups. This means, across 30 teams, the average shortstop would record 510.16 outs in 2013.

We’ll be looking to see if Simmons was able to record 510.2 + 105.8 = 616 outs in 2013 to merit his 5.5 dWAR.

It turns out, in fact, that is the exact number of outs he recorded (499 assists + 117 batted-ball putouts = 616 outs) on batted-balls.

In 2013, Simmons prevented 106 more hits than an average shortstop which roughly equated to 5.5 wins that season.

I believe the error in Peta’s math comes from the 2013 total of 13,772 assists and 3,219 batted ball putouts he uses. Perhaps these figures included playoffs or maybe they counted the total batted-ball outs recorded by any player who played shortstop at all in 2013, instead of when they were specifically playing shortstop, or maybe there is some other source of error, but according to play index, the 15,305 figure is the accurate combined total for 2013 shortstops.

Now, I still believe some external factors make his 2013 campaign even more impressive.

The 2013 Braves pitching staff faced the fewest number of batters (5,989) in all of baseball and allowed the 9th fewest balls in play (4,221). Simmons was able to distance himself 106 outs above the average shortstop despite facing less than the average number of opportunities.

To be fair – when the ball was put in play against the 2013 Braves, it was sent in Simmons’ direction a disproportionate amount. The Braves’ pitching staff carried the 6th highest GB% in 2013. However, I think it’s a mistake to ding him for this.

First, while ground-ball induction is a pitching skill, it’s also a factor of location and game-planning. With a cerebral catcher like Brian McCann, I think it’s more than conceivable the Braves tried to induce ground-balls to Simmons whenever possible because of his excellence. Penalizing him for this is a bit like saying “Of course Tiger Woods has won more WGC’s than anyone, he’s had the most opportunities.” Woods has had the most opportunities because he’s the best in the world at what he does and earned his way into WGC’s – counting metrics reflect this. Simmons earned the right to have ground balls (at least very slightly) directed his way, because he’s the best in the world at what he does.

I think we all know who the dot in the upper righthand corner is, so I won’t label it.

Second, increased opportunity does not automatically translate to defensive performance (and if it does, it should be appreciated, not discounted).

In 2013, Defensive Chances (Putouts + Assists + Errors) had an extremely low correlation to Defensive Runs Saved above average among shortstops. Of course a counting stat will reward players with more opportunities – this is how it should be.

Reducing Simmons to being nothing but an elite “hit preventer” is also overly restrictive. The raw counting of defended hits overlooks the context under which they occurred – both by leverage index in the game and degree of difficulty on a specific play – and neglects the runs he saves by being the best double play infielder on Earth.

One of the biggest holes in Mark Broadie‘s  tremendous work with Strokes Gained in golf is that not all 8-foot putts are created equal. Strokes gained does not account for the fact that a down-hill, left-to-right slider to save par and force a playoff on the 72nd hole of The Masters is materially more difficult to make than a dead-straight, flat 8-footer on Thursday of the John Deere Classic. This isn’t a huge flaw, because for most golfers, with enough of a sample size, these extremes will roughly cancel one another out to avoid any huge mischaracterization of talent. However, there are some players who are particularly good at course management and positioning themselves to be more “beneath the hole” than average pros. For these players, strokes gained may underestimate a player’s putting ability because a higher than average percentage of his putts are downhill or breaking due to his poor course management. The player deserves to be docked – no doubt – but not specifically his “strokes gained putting” metric.

In baseball, the same issue arrises when we only look at raw totals of “plays made above average.” Another reason Andrelton Simmons is repeatedly near the league lead in Defensive Chances is his freakish range to create an above-average number of “fringe” or “high degree of difficulty” plays that many shortstops don’t even have the ability to get a glove on. Because of this, it’s possible Simmons’ defensive stats are actually underpricing his defensive value. To relate back to strokes gained – Simmons is repeatedly facing an above-average percentage of downhill breaking putts, and still makes them more at a more than league-average rate (the only difference is this obstacle isn’t born from poor course management, but rather from the gift of being more athletic than his peers).

Simmons’ separation from the rest of the MLB even causes a slight negative bias in stats that are based on league averages. He is such an outlier and so far ahead of second place in so many categories that he causes a severe right-skewed distribution of most defensive stats every year. Using a metric like “Defensive Wins Above Median-Performance” would make his results even more staggering because it would remove his greatness from the sample of data that is raising the standard of “average” up.

As is usually the case when two intelligent people are diametrically opposed – the truth is probably somewhere in the middle. I’ll let Ozzie stay within breathing distance of Simmons on the all-time defensive leaderboard if Joe takes back his “homeless man” statement.

Andrelton Simmons’ consistent dominance of all aspects of MLB fielding fascinates me to no end. He has remained so steadily head-and-shoulders above his peers, in nearly every category, for a remarkable length of time and with no clear end in sight. As a bettor, my primary focus in evaluating, analyzing and most importantly – projecting – a player is to get a good estimate of what his presence or absence means in a single 9-inning game of baseball. My Simba obsession goes beyond just the fun and quirky nature of his greatness – it’s born from anecdotal and scientific betting success from valuing his absences. There are many flaws with the stats I’m about to provide but…

MLB teams record when Andrelton Simmons plays shortstop for them: 437-443 (.497, ~81-win team)

MLB teams record when Andrelton Simmons is out of the lineup for them: 39-53 (.423, ~69-win team)

Again I’m sort of cringing by sharing this because it’s only slightly better than meaningless, but being able to sufficiently value his absence – especially in 2016 – has helped me and my bankroll a lot.

More entertaining than any of those units, though, is the sheer quality and quantity of highlights he’s given baseball fans the last six seasons. And he hasn’t even “made the turn” yet…

 

 

 


Kieran

is a bitcoin developer and professional sports bettor.

leave a comment: