Diamondbacks Should Bid $20M or More for A.J. Burnett

A.J. Burnett was an ace last season. In 2013, he was eighth among qualified starters in FIP (2.80), 21st in WAR (4.0), and 27th in ERA (3.30). But his 2013 performance is not the only thing that makes him a good bet for the D-backs — Burnett was outstanding in ways that happen to be highly predictive for 2014.

It’s true that projection systems don’t count on a repeat performance. ZiPS projects Burnett at 3.50 FIP, 3.58 ERA. Steamer projects Burnett at 3.27 FIP and 3.63 ERA, good for a 3.4 WAR in 189 innings. That’s a cut above Ubaldo Jimenez (Steamer projects 3.79 FIP, 4.02 ERA, 2.5 WAR) and Ervin Santana (3.71 FIP, 3.95 ERA, 2.7 WAR). Burnett has also been extremely durable for quite a while now, pitching at least 186 innings in the last six seasons (1,198.2 IP in that span). I know, he doesn’t have that reputation. But the numbers don’t lie, and there are at least two reasons to think he’s even a better bet than the projections imply.

It’s the sinker

The first is that, while both projection systems take an enormous amount of information into account, they don’t necessarily know the principal explanation for Burnett’s resurgence: his new(ish) sinker. At Beyond the Box Score, Jake Dal Porto wrote that after he was traded to the Pirates before the 2012 season, Burnett started using a sinker much more than he had previously, and much more than his 4-seam fastball. That brought Burnett’s GB% up quite a bit, and then it was off to the races.

I can’t be the only one that trusts recent success more when there’s a plausible reason for it, as when Alfonso Soriano starting hitting bombs all over the place after switching to a lighter bat. Fixing an eye problem, as with Freddie Freeman. Changing up a batting stance, as with Brandon Belt last year. Adopting a new, great pitch that has great results? That might be the most trustworthy of tidbits overall.

In short, there’s a reason to think that Burnett’s last two seasons of work have a disproportionately high bearing on what’s likely for 2014, because his pre-2012 track record may have less bearing on his 2014 than is generally the case for starting pitchers. You don’t have to agree with me that ZiPS and Steamer are at least a little pessimistic, but maybe you’ll at least concede that they’re not overly optimistic.

The K% and GB% for Burnett are great signs

The other big reason to trust Burnett’s recent success may be the most interesting: he was good in a way that’s highly predictive of future performance. Defense-Independent Pitching Statistics (DIPS) theory is based on the premise that there are some things that pitchers can control, and some things they can’t. Walk rate is one thing that’s completely independent of fielding. Strikeout rate is another, and ground ball rate also appears to be a repeatable skill.

The reason why ground ball rate is helpful, in addition to being stable, is that it’s pretty damned hard to hit a home run on a ball that bounces before leaving the infield. It’s really just a matter of home runs: as an event, non-HR fly balls have a run value of -0.12 (see comment 8 here), equal to the run value of ground balls. Once you put HRs back in the mix, the run value for fly balls jumps to 0.03. That’s why we’ve heard so much about the Athletics’ latest strategy.

I don’t need to explain that strikeouts are good things for pitchers (after all, you’ve shown yourself to be smart just by choosing to read this, right? Right?). But I don’t know if it’s just as obvious that it’s so, so difficult to have strong strikeout rates while posting better than average ground ball rates.

Ground ballers like Brad Ziegler tend to have high contact rates, and high contact rates tend to mean low swing and miss percentage, and low swing and miss percentage tends to mean low strikeout rates. As pitching statistics go, K% and GB% have a weak, but inverse relationship in the same season (-0.233 at minimum 50 IP), courtesy of Steve Staude’s correlation tool at FanGraphs). It’s hard to buck that trend.

But that’s just what Burnett did in 2013, better than anyone in baseball. As Bryan Grosnick found, when the extent to which a pitcher had an above-average K% was slapped together with the extent to which he had an above-average GB%, Burnett ruled the world in 2013 as baseball’s “Winter Soldier.”

Ok, so why should we care? As Grosnick observed, K% has a high correlation year to year (0.702), and GB% is even more stable (0.752) (both numbers with a 30 IP minimum). Among “elite” GB% starters, the correlation continues despite the fact that such pitchers are outliers (0.684, min 100 IP, min 50% GB%, n=117).

Burnett’s value

Pitching enough innings to have enough to qualify is almost definitely a sign that a pitcher is worth something, so narrowing the field of 2013 starters to 81 is already putting all 81 of those pitchers in good company. But by FanGraphs’ “Dollars” metric, Burnett ranked 21 among those 81 pitchers, at a striking $19.9M. The rate of dollars per wins on the open market has gone up steadily (it looks to be north of $6M/win this offseason), but the progression is linear, as teams don’t pay exponentially more for higher win value players.

Last year’s $19.9M “Dollars” figure is not the end of our analysis. I think Burnett is a strong bet to beat his projections for next year, but if we use Steamer’s 3.4 WAR projection, Burnett will fall short of $19.9M “Dollars” in 2014 (which was based on 4.0 WAR). Not all the way to $17M “Dollars,” mind you — the going rate for a win is definitely higher this offseason than it was last offseason, which could bump a 3.4 WAR performance back up to the $18.5M range.

So why bid higher than $18.5M? I can think of three reasons (and maybe you can think of more). One reason is that there’s less risk in a one-year deal, and it seems like that’s all Burnett will require. If Burnett would require a $18.5M AAV for a longer deal, there’s some reason to believe, I think, that he’d be worth more than that to the D-backs on a one-year pact, which could easily take the amount above $20M. Another reason is that the D-backs may actually be motivated to pay exponentially more for this higher win value player, because they’re looking to improve from a starter who actually could be pretty decent in 2014 in Randall Delgado.

But the biggest reason of all may be that it could be tough to convince Burnett to come to Arizona. Of course, the fact that Burnett might need to be offered more money to come to Arizona does not mean that the D-backs should pay that amount. But at this point in the offseason, and with the D-backs already making some bets on 2014, why wouldn’t they pay a premium, if that’s what was necessary? This is not a matter of a three year deal, when paying a $2M-$4M premium on an AAV basis could cost $10M. It’s not a matter of paying a first-year arb player an extra $2M, which could also end up costing the team $10M overall. This is a single shot. Overpaying, even to the tune of a $22M deal, will not have lasting consequences.

Reasonable minds can differ on whether the D-backs should make an offer to Burnett, and on the amount of any offer that is made. But I think a strong offer in the $20M-$22M range is a smart idea for the club. In this year, it makes sense for the team to make one more impact move. And among these starting pitching options, Burnett offers both a high level of potential reward and a low level of risk.

For more on why A.J. Burnett may be the best pitcher available on the market right now, check out this article by Alex Skillin.