Behind Dombrowski’s Love of Velocity

facebooktwitterreddit

We at MCB have been fairly hard on Dave Dombrowski regarding his bullpen personnel strategy this offseason – as I suppose has the national baseball media as well. The Tigers have shown up on “everybody’s” lists of potential suitors for this free agent reliever and that one, but without so much as a peep from Dombrowski himself who has assured us all that the Tigers are done shopping. We shouldn’t expect any noteworthy changes to be made – and come to terms with the 2014 relief roster we have generally panned.

Aug 7, 2013; Cleveland, OH, USA; Detroit Tigers relief pitcher Bruce Rondon (43) reacts in the ninth inning against the Cleveland Indians at Progressive Field. Mandatory Credit: David Richard-USA TODAY Sports

As I have mentioned before, what seems to be behind this offseason’s bullpen decisions as well as those in the past is a strong preference – an organizational preference at all levels – for guys who throw hard. The Tigers 2014 ‘pen will be down two of it’s softer throwing regulars in Darin Downs and Drew Smyly as well as a moderately hard thrower (by comparison to the Tigers other options) in Joaquin Benoit. Their replacements (Ian Krol, Joba Chamberlain and Joe Nathan) throw harder as a group – we should expect that the Tigers 2014 bullpen will throw a bit harder than the 2013 group on average even if we think of the unit as an organizational one that goes 12 deep. Though the guys that they are replacing were darned good (though Downs only in terms of peripheral stats) the added velocity isn’t something that we should pooh-pooh.

Every baseball fan knows that throwing hard is better, but they may not be familiar with just how much better it is – statistically speaking. I’ll try to explain what exactly throwing harder “does” for the ‘pen in terms of game outcomes using a statistical measure called a “correlation coefficient” (calculated at the team level for every team-season over the past decade). This measures how strongly two things tend to go together, scaled as the percentage of the total variation in one statistic that appears to be due to it moving in lockstep with another things. Everybody knows that throwing harder leads to more strikeouts – right? The “correlation coefficient” between velocity and strikeout percentage is 45.3% – 45.3% of the variation in strikeout percentage tracks variation in velocity, the remaining 54.7% is due to other factors or blind luck.

That’s a pretty strong correlation, the number between ERA and xFIP is only 70% – with the remainder made up (if you’re a believer in the advanced statistics) of some combination of smoke & mirrors and dumb luck. The correlation between velocity and xFIP or SIERA (the “composite ERA”) is about the same as the correlation between velocity and strikeouts: 41.1% and 41.2% respectively. The correlation with ERA is a still-high 29.5%. The higher correlation with xFIP and SIERA isn’t all that surprising as these stats are driven in large part by strikeouts – but that isn’t all there is to it. There seem to be smaller but still “beneficial” correlations between velocity and a whole host of other game outcomes.

One thing that extra velocity does not do is make a guy throw strikes – the correlation between velocity and “walk percentage” is tiny but positive, meaning that more velocity goes hand in hand with a higher walk percentage. BUT… a higher velocity is actually correlated with fewer walks per 9 innings (a correlation coefficient of about 4%). The reason that is possible is that guys that throw harder tend to see fewer batters per inning. The correlation between velocity and WHIP is a negative 26%. So what do we get in addition to the extra strikeout here and there? We get a lower BABIP – a correlation coefficient of negative 7.4%. That’s statistical mumbo-jumbo, but just another way of saying that it seems that when a guy throws harder the batter does appear to make at least slightly worse contact. This can also be seen in a correlation coefficient of negative 10.5% between velocity and home run to fly ball ratio and a correlation coefficient of 8.9% between velocity and ground ball rate. The last key thing to point out is that there is a strong correlation between strand rate (LOB%) and velocity at 23.8%. Bullpens that throw harder allow fewer guys to reach base in the first place and they allow fewer of those that do reach to score.

Everyone also knows that not all guys that throw hard throw well – but it’s easy to understand why a tangible attribute that seems so closely related to the ingredients of success would be what Dave Dombrowski pursues above all else. The Tigers had the 6th hardest throwing ‘pen in baseball last year and are likely to move up. The hardest throwing ‘pen was the Royals – who were also almost unarguably the best. The Tigers peripherals were vastly better than their ERA last season – as was also the case for another hard-throwing but mediocre unit in Seattle. If you’re looking for a reason for optimism about next year’s bullpen you can bank on a bullpen that’s top-5 in velocity at least falling in the top half in terms of strand rate (instead of 23rd) and BABIP (instead of 24th) and HR/FB% (instead of 17th).