The Math Behind Jon Lester’s Harmless Oddities
Jeff Sullivan wrote a post this morning about Jon Lester and the running game. He mentioned that I’d also be writing a post about Jon Lester and the running game, but with a greater emphasis on the numbers side of it. This is that post.
Before we get to the actual numbers, a note about Jon Lester himself. In a way, for much of his career, Lester almost been consistent to a fault. To the point where his greatness borders on boring, or forgettable. In nine years since taking on a full workload, he’s made between 31 and 33 starts in each season, always 191 and 219 innings. He had a three-year run of his ERA- being 71, then 73, then 75, and four of his nine FIP- have been between 73-76. His fastball has sat between 91.8 and 93.5 miles per hour — right at or below average — in each of those nine years. Lester’s had his two best seasons by ERA in the last three years, but even then, his FIP- figures have read: 75, 75, 82. Just consistent ol’ Jon Lester. Nothing remarkable here.
And yet, somehow, the longer Lester remains consistent, the more we realize he’s one of the most fascinating and unique specimens in the game. We realize he simply refuses to attempt a pickoff throw to first base, and that’s because when he’s forced to field a ground ball and make an overhand throw to first, he just literally can’t do it. The pitcher just cannot throw. We realize that he’s maybe the worst hitter, ever, like in MLB history. And so we watch each one of his starts with amazement, as the gifted, elite athlete is unable to hide his inexplicable ineptitudes, and as the opposition just… fails to exploit them?
Give the Dodgers credit. They sure as hell tried. Kind of. At the very least, they sure as hell put put all of Lester’s bizarre quirks front and center stage in their 8-4 NLCS Game 5 loss on Thursday night. It’s just, none of it mattered.
The Dodgers wasted no time letting Lester know that they knew. This was the first pitch Lester threw:
That’s Dodgers leadoff man and left fielder, Enrique Hernandez, feigning bunt while Lester threw a first-pitch ball. I’m not sure whether Hernandez actually intended to ever lay down a bunt in this at-bat, but the thinking goes: you show bunt, you make Lester think about fielding, and then, subsequently, throwing, and you’ve gotten in his head, therefore making him give something less than his full attention to executing the actual pitch at hand. Who knows how much validity that line of thinking has, but, hey, Hernandez showed bunt on all four pitches, and Lester threw balls on all four. Hernandez found himself on first base.
And then he found himself getting gradually farther from first base:
And he danced:
And he confidently toyed with Lester:
And he pretended to steal:
And… he never actually stole. Hernandez did all his dancing and strutting and toying and letting everyone in the park know that he knew about Lester’s big, embarrassing flaw, while Lester struck out Justin Turner and Hernandez stood on first base.
Hernandez eventually advanced. He advanced exactly one base on a single, clear as day off the bat, by Corey Seager:
One way or another, it’s hard to fathom how Hernandez isn’t on third base here. Despite the fact that Lester has attempted precisely three pickoff throws since April 13 of last year and hasn’t even attempted a single pickoff since July 24 of this year, Hernandez takes a several steps back toward first after getting his lead out to about 20 feet. You see his body when the pitch is thrown; he gets no secondary lead, has no momentum.
And the Dodgers continued this trend all game. Here’s Seager in the sixth, his momentum taking him back toward first while the pitch is thrown:
On the surface, it’s baffling. Why take the lead in the first place if you’re (a) not going to do anything with it, and (b) are just going to retreat when the pitch is thrown, killing both momentum and your distance to second base? Well, the latter has everything to do with catcher David Ross. Ross loves to throw down to first, and he’s got the arm to get the outs, and so the runners have to respect it. OK, so then why take this big lead in the first place? Why not either take a normal lead with a gigantic secondary, knowing you’ve got no change of being caught in the middle of a deceptive Lester pickoff, or just get a running head start and take the bag?
Well, let’s run some math. The problem here is, Lester and Ross are quick. All of the following information comes from Statcast, provided by Mike Petriello. Lester was getting the ball to the plate between 1.1 and 1.2 seconds last night, and Ross’s average pop time for the season was 1.95, which ranked sixth among 83 catchers with at least five throws to second base. Both those figures could be considered plus to elite on a scouting scale, and as a battery, their ~3.15 time to second base is hard to beat.
Now, Hernandez has only ever stolen two career bases, and one was a botched pickoff attempt, so there’s not much in the way of speed time data for him. But the league average, on successful steals from first to second, is around 3.7, and the league-average lead is 11 feet, meaning our hypothetical league-average runner is covering the remaining 79 feet in those 3.7 seconds, putting him at 21.4 feet per second.
But Hernandez wasn’t getting a league-average lead. He was getting up to 17 feet on Lester. That leaves just 73 feet, and, giving him league-average speed, that gets him to second in… 3.4 seconds. Lester and Ross have him beat by more than two-tenths of a second. Even a league-average runner with a 20-foot lead is getting beat by more than a tenth. But maybe Hernandez — or any other base runner, really, Hernandez is just a placeholder here — is faster than league-average. Someone like Jarrod Dyson can go first-to-second as fast as 3.0 seconds, so let’s put our guy right in between, at 3.4 seconds with an average, 11-foot lead. Ramp that lead up to the 17 feet Dodgers runners were taking on Thursday, and that gets us to second in… 3.15 seconds. Exactly Lester and Ross’s time.
A quick pitcher and a quick catcher can go a long way. Runners may have stolen 28 bases while Lester was on the mound this year, the third-highest total allowed by any pitcher, but they were also thrown out 13 times, and a 68% success rate actually reflects well on Lester and Ross.
What about the bunts? It wasn’t just Hernandez who led off the game by showing bunt. The Dodgers showed bunt against Lester more than a dozen times. Adrian Gonzalez bunted out to a second baseman who started the play in the outfield grass in the seventh. Joc Pederson missed a bunt to go down 0-2 in the count in the fifth inning, and then hit the ball 110 mph on the next pitch. And this came after Pederson actually laid down a bunt in the second:
On the one hand, I mean, you see that throw. On the other, apparently that’s something Lester actually worked on this year, and when you look at the play logs, you find that Lester’s fielded 20 ground balls or bunts this year, recorded 19 outs, and committed zero throwing errors. At a certain point, it started to feel like some of the Dodgers’ best hitters were taking the bat out of their own hands.
Even though another one of the bunts Lester fielded this year went like this…
… and looked like a problem, it wasn’t actually a problem, just like Pederson’s bunt wasn’t actually a problem — just like Lester’s inability to as much as even attempt a pickoff throw sure looks like a problem, until it just isn’t. None of us will ever truly understand any of this, but that’s how it goes.
Greg Maddux never held baserunners, either. He allowed 547 steals in his career, and up to as many as 37 in a single season. He just didn’t care, because he was good. Over the last three years, Lester’s allowed one of the lowest rates of baserunners in the game. He and Ross are exceptionally quick in their deliveries and exchanges, making it difficult for even above-average baserunners with gigantic leads to steal successfully. Hesitate on the steal, and Ross might nab you at first. Try to challenge Lester with a bunt, and, well, bunting is hard, and he might field it in a position to just underhand it to first, anyway. The thing about trying to get into Jon Lester’s head is, you spend too much time trying to do it, and you might wind up getting in your own.
August used to cover the Indians for MLB and ohio.com, but now he's here and thinks writing these in the third person is weird. So you can reach me on Twitter @AugustFG_ or e-mail at august.fagerstrom@fangraphs.com.
It sure sort of felt like the Dodgers were in their own head more than they were in Lester’s last night. The Gonzalez bunt in the 7th, especially, was trying too hard imo.
i think that one wasn’t just about jon but how deep baez was being played to Gonzalez he was near 200ft out i recall the broadcast showing. Plus i think they were hoping a man on in the 7th may chase Jon
Yes, I agree with Otter that the Dodgers seemed to be in their own head, and yes, Gonzalez was out on the bunt, so you could say it was de facto the “wrong” gamble. But it took a spectacular play from Baez to get him. On that bunt he may be safe – what – 40% of the time? Those are pretty good OBP odds – perhaps enough to justify his decision to bunt, even if the outcome didn’t work out.
I wonder of that 68% success rate how many are because of true burners like Hamilton and Villar. Against “normal” basestealers i bet 68% reflects worse than they really are
obviously i’m aware that’s true for everyone but seems more pertinent
when the pool of runners is limited
I think it reflects runners not realizing their lead can be over 20 feet more than anything. 20 ft lead plus momentum going to second plus going on first move should be 100% success.
Yes, and what is the practical limit on how much of a lead the runner could take on Lester if said runner is committed to stealing, going first movement on first pitch (so no pickoff throw from the catcher)? Could 20 feet instead become even 25 feet or 30 feet? At some point, I suppose Lester would step off and just run at the baserunner until getting close enough to a base to throw the ball underhanded?
It’s admittedly tough for a team to experiment with pushing these tactics really far in the playoffs, but there’s more upside for a division rival that can expect to face Lester roughly 4 times per year for multiple seasons.
I am just repeating what someone else said, but I think it makes sense. If a runner gets 30′ off the base, that is 1/3 of the way to 2B, couldn’t Lester just step off the mound and run right at him, probably just underhanding a throw to 1B to get him out?
Stepping off, running toward runner plus Underhand throw plus tag is really really slow though. 30 feet is most likely safe to be honest. Over that you are probably pushing it though.
Oops. I put this in the wrong reply. You are the one who said it already. Good job.
Are pickoffs on throws to bases by Ross reflected in that 68% number?
I could watch that last gif for hours. That is what I would look like if I tried to make that throw with my left hand, too.
Apparently what you’ve been missing is throwing off a mound?
That’s all well and good, but we still can’t answer your own question: “Why take the lead in the first place if you’re (a) not going to do anything with it, and (b) are just going to retreat when the pitch is thrown, killing both momentum and your distance to second base?”
If the Dodgers know they can’t really exploit Lester and Ross the same way that the Royals exploited Lester and Norris in the 2014 AL wild card game, for example, then way do things that hurt your chances to go from 1st to 3rd? That run probably scores in the first if Hernandez gets to 3rd.
Lots of people seem to be falling all over themselves praising Dave Roberts’ managing and thinking, but IMO, he’s taken things a step too far. Taking out Maeda with Lester coming up was pure insanity. He was burning through RPs who were facing 2-3 batters in the 4th and 5th innings. But then in the 7th as Baez was imploding he was allowed to face 7 batters and got 2 outs. And its not like Baez is some lights out RP that deserves to be trusted, he’s got a FIP of 3.81 this year… just what the ef….
To the last part. Over small samples (74 innings) it’s better to use xfip because homer rate doesn’t stabilize that quick. Not too much better but Baez was solid. I agree on meada but lester did square him up in the first ab lol.
Roberts’s RP usage definitely questionable. Relying on Baez and Blanton in 8th inning of 2 close games, while holding back Jansen, MIGHT be defensible if they were really proven as good as their season stats. But both had much higher workload this year than ever before *as relievers*, and showed the effects in the 2nd half — Baez OPS up 94 pts, Blanton up over 200.
LA tied up game 1 in top 8th off Chapman, the Cubs’ 5th RP used. Dodgers were about to use their 3rd RP, so they seemed in better shape for possible extra innings, and IMO a great time to go to Jansen for 2 innings — or at least use his massive K rate after the leadoff double. Didn’t happen.
In game 5, LA trailed by 2 into the 8th — okay, not the most obvious time to use your relief ace. But Jansen didn’t pitch the day before, a day off was coming up, and the odds of a save opp were getting slim — why in hell *wouldn’t* you use him? They had the top of their order due in the bottom half, and the Cubs’ late relief hasn’t been dominant. I say you’ve gotta sell out to stay close there — if not start the frame with Jansen, at least use his massive K rate after Cubs got to 3rd and 2nd with 1 out. But, nope.
Game 1 of the NLCS was October 15. Jansen had thrown 51 pitches over 2.1 innings on October 13 in Game 5 of the NLDS. My recollection is that Roberts had said that Jansen would be available in Game 1 of the NLCS, but for no more than 1 inning. Jansen definitely wasn’t going to pitch 2 innings in that game, and I took the idea of 1 inning from him with a grain of salt. It’s certainly fair to wonder how effective Jansen would have been pitching in Game 1 after such a heavy workload just 2 days prior.
In Game 5, I can see the logic of starting the 8th with Baez. The Cubs’ 7-8-9 hitters were due up, so Baez for the 8th followed by Jansen for the top of the order in the 9th was a logical way to try to keep the Cubs from adding to their 2 run lead. As you say, though, once it got to 2nd and 3rd with 1 out and the top of the Cubs’ order coming up, good idea to use Jansen to try to prevent any runs from scoring.
The playoffs are not a time to start experimenting, but I don’t understand why teams have not tried to take bigger leads or even run to 2nd, especially teams that have
nothing to lose. If you get picked off a few times then you know it doesn’t work.
I wonder if the risk-reward calculation just argues against it. In the regular season whatever marginal advantage you might gain by trying to run wild on Lester may not be worth the hassle of deviating from your usual game plan. If you’re not already a good base-stealer, or a team that steals as part of its offensive game plan, there may not be much value in trying to “break” Lester. Especially since this article indicates that the risks of trying to do that are actually a lot more signifiant than they seem from a fan’s perspective. And the same calculus then applies in the playoffs–if you successfully run all over Lester then that’s very valuable, but if you go outside your comfort zone the risks are significant as well. The team that seems best-suited to exploiting Lester is Cleveland, because they actually run well. They’d be exploiting Lester’s weakness with their own strength, not hoping that Lester’s weakness is so severe that it turns one of their own weaknesses into an advantage.
I agree that it’s tough for a team to go outside its comfort zone in the playoffs.
For a division rival in the regular season, however, I see a lot more potential upside. Division rivals play each other 19 teams per season, so the expectation is to face Lester about 4 teams per season as long as he’s healthy and with the Cubs. That’s decent upside if a team hits on a tactic that works.
What if, for example, a team decided to take absolutely massive leads at first against Lester – like 25 or even 30 feet – and then try to steal on first movement on first pitch every time (so Ross can’t attempt a pickoff throw at first)? If it doesn’t work, you would decrease your chance of winning one game. If it does work, though, you probably have a tactic that helps your win probability in several games per season for the next few years (unless Lester figures out how to throw to first or otherwise adapt).
Yeah, “what can we do to take advantage of Lester?” seems like a question that ought to matter most to the Cardinals (and maybe the Pirates, if they bounce back next season). When it’s a division rival AND a team you might face in the playoffs, the downside to trying something that doesn’t work in a couple regular season games is probably outweighed by how valuable it would be to hit on a workable strategy.
I agree. The risk-reward calculation might be for not running. I don’t see why more teams have not tested it. It seems similar to mangers only using their closers in the 9th inning.
I think mangers have more incentive to look for ways to maximize the effectiveness of their bullpens, because the successful strategies are going to be so valuable so often. In the regular season, even for other NL Central teams, there’s just not that much advantage to be gained over the course of 162 games by trying to perfect your Lester-killing base-running strategy. Especially if, like the article says, it’s actually tougher and riskier to grind out that marginal advantage than it seems.
I keep coming back to Game One of the NLDS as I think about this. The Giants clearly tried to take advantage of Lester’s throwing issues. In the first Gorkys Hernandez bunted for a base hit. He then tried to steal and was caught (making him 8-for-12 in career MLB steal attempts). Gillaspie got on base to start the third inning and was picked off by Ross and Baez while Rizzo guarded against a bunt; Cueto then bunted with the bases empty to make the second out.
I’m not listing all that to say that the Giants were reckless or pursued a bad strategy. It seems like a good strategy to me. But as it played out it ended up running the Giants out of two innings in a game they eventually lost 1-0. That indicates to me that managers may not be wrong if they decide to put together a strategy that weighs their own team’s base-running ability more heavily than Lester’s throwing.
Willson Contreras has clearly been learning from Ross; his snap back throw to first has improved and he’s “stabbing” at the pitch less than earlier in the season. Hopefully Lester & Willson can get comfortable together next year after Ross retires.
i’d hope even with ross retiring they may have a 3 man boot camp planned over the offseason
Lester should just bounce the pick-off the same way he does the regular fielding plays. Runners would be toast.
You’re assuming that Lester can make that same bounce throw just as well on a pick-off, which very well might not be the case. Lester obviously has a mental block, and it’s not crazy to suppose that standing on the mound with time to think about a pick-off throw is worse for him than fielding a ball and making a quick throw.
My recollection of the throwing yips of both Steve Sax and Chuck Knoblauch, for example, was that they showed up on the easiest, most routine ground balls with plenty of time to think about the throw to first. When Mackey Sasser developed problems throwing the ball back to the pitcher, he was still fine throwing to bases on stolen base attempts.
I had the yips as a shortstop for about a month. It was due to a runner beating out a grounder I was lackadaisical on. It caused me to start rushing routine plays, but my cutoffs from 200 feet would still be in the money and I could make the diving plays because instincts and athleticism took over. It was when I told myself to make a good throw I didn’t.
So I wonder if that’s, say, 0.6 seconds better than the average battery, how many feet of a lead off first base does that equate to? That would be a really cool use of statcast. A 15 foot lead on Lester/Ross is the same as a … 9 foot lead on Arrieta/Montero?
The thing with Lester is you can not only get a huge lead, you can literally go into a near sprinters pose and he won’t throw over. And if you are going to get that sort of lead, you should be stealing on first movement even if he throws over so all the momentum is going towards second. That’s not a standard 17 foot lead for an average pitcher. When the runner has committed himself to going to second on anything beyond Lester stepping off and running towards him, nothing is ordinary.
So what’s Lester’s and Rizzo’s combined time to second?
This is the whole issue. Because nobody actually runs, there isn’t any theoretical advantage to the runners (especially with Ross holding their leads down). The thing with Lester isn’t “big leads” but “big leads that become stolen bases.” Those haven’t materialized. Hernandez got 20 feet from Lester… just start running before he throws home! Nobody has tried that, they’re all too scared or hesitant, but until Lester proves he and Ross can handle that (which they haven’t, nobody has tried to make them prove it), GO!
Note: Rizzo had a 20 steal season despite his lack of speed… because he took advantage of pitchers not throwing over! Yet, nobody seems to be able to do this with Lester. They can’t get out of their own heads…
There are a lot of things in baseball that people would say are “fascinating”, but really they’re “just” interesting. Not this. This is legitimately fascinating.
What’s remarkable about how awkward Lester is to 1st base, be it a pickoff throw or just ordinary throw, is that Lester didn’t look at ALL like this early in his career. In fact, early on I thought he had the making of a good pickoff move. I’m not sure where it went wrong, but it’s a credit to him that he hasn’t let it snowball into disaster even as he isn’t able to outright repair the flaw.
I doubt any of us will fully understand how difficult pitching at a major league level is. (or hitting for that matter, but we’re talking about pitching) I don’t begrudge Lester his quirks, though there’s no question he’d be a more complete pitcher (in terms of DRA) if he had a pickoff move… any pickoff move.
The Maddux anecdote is really telling: the best way to prevent steals is to prevent base runners in the first place. Lester isn’t in Maddux’s league, but a 70-80 FIP- is still elite. Pair that with an elite defensive catcher and a naturally quick delivery and the problem is basically erased.
How much faster is Ross than the average catcher? Will his retirement be a big issue for Lester?
Contreras is quick too with an even better arm.
The Royals stole 7 bases off Lester in the 2014 wc game…
Without a catcher with a plus pop-time and before his yips were common knowledge. While I still worry that a team that steals bases often and efficiently (like, say, the Indians, gulp…) can take advantage of those leads more effectively than the Dodgers or the Giants, I have doubts that it’d amount to the same disaster that was the 2014 wild card game.
Still, it seems to be that even with .2 seconds for Lester and Ross to beat you to with the throw, it’ll still probably take a near perfect throw to get an even slightly better than average runner out. Why? Because it takes time to bring the ball down on a high throw, and a head first slide could have the runner sliding wide and tagging the base with their left hand, or trying some other evasive technique. If you can get a 17 ft lead off of first and have any speed at all, it might still be worth it to test the two, if you get lucky and correctly call a Lester breaking ball or off speed pitch, which he’s thrown close to 20% of the time this year, you can expect slightly longer time for Ross to be able to receive and pop up with the pitch (not to mention if it goes in the dirt, behind the batter, etc.
If the Cubs advance, I fully expect the Indians to test the pair with Lindor, Kipnis, Davis, and J-Ram whenever possible.
Not only does Lester have a quick delivery to the plate, he also does an incredible job of varying when he starts his delivery. Sometimes he will hold the ball in the set position for a long time, other times he will throw it quickly.
Lester also varies where he looks as he starts his delivery. Sometimes he will be looking at the catcher, other times he will be looking at the runner at first even as he starts his delivery. Sometimes he will look back and forth.
What this all means is that even though it is easy to get a good lead on Lester, it is incredibly hard to get a good jump on Lester. Bad jumps combined with Ross’s good arm leads to lots of caught stealing attempts.
While it is easy to say “Go on first movement”, that is much harder to do in practice when first movement can vary between 1 to 5 seconds and there are no visual clues such as where he is looking. Combine that with the fact that Lester steps off the rubber just often enough that going on first movement means he steps off the rubber, jogs towards second, underhands the ball to the 2nd basemen and you are out by 5 feet.
Good jumps are far more important than good leads when stealing bases and Lester is one of the hardest pitchers in baseball at getting a good jump on.
Since he doesn’t ever throw over, once his foot comes up he’ll pretty much have to throw to the plate, that’s when you go.