Charging the Mound: Ride the Hot Streak?

Subject: Charging
From: Christopher Liss
Date: August 7, 2013 10:43:10 PM PDT
To: Andrew Martinez

I think this might be my favorite part of the baseball season - when your teams (good or bad) are what they are, you don't have much FAAB left, most of the big trades to be made have been made and you can't do much but sit and watch. Let's see if Giancarlo Stanton and Bryce Harper reach 30 HR - forget that they're behind that pace, either is capable of hitting 15 HR in a given month. It's less oppressive than early in the year when you have a million decisions, and many of them are for the long haul. This is when it gets a little easier (until the September call-ups, when all hell breaks loose again).

Even for my teams that are out of contention I'm watching to see how my guys perform so I can get a sense of whether I was actually right about them after all. If you spent good money for B.J. Upton, and he hits .290 with 10 HR and 14 SB the rest of the way, you won't second guess yourself quite as much during the offseason. He'll still have sunk your team, but you'll realize it wasn't as if the skills had disappeared, and you were too blind to notice. It was due to something else in that case. But if the guy has a disaster season all the way through - Adam Dunn in 2011 comes to mind. (I still remember keeping him active and feeling sure the moment I benched him he'd hit 10 homers in a week.) Guys I'm watching to see who they are include Jason Heyward, Starlin Castro, Brett Lawrie, Ike Davis, Jurickson Profar and Devin Mesoraco. With pitchers, you never know who they are, so there's no point. They're someone for a while, and then it wears off, and they become someone else. Only a rare few can sustain any sort of identity over time. Maybe Matt Harvey and Jose Fernandez will be two of those guys, but remember Tim Lincecum won two CY Youngs, and he's a shell of his former self at age 29, his last couple good starts notwithstanding.

One topic that came up on the radio last week - and we deal with it all the time in daily games - is whether a player's recent hot or cold streak is an indication of whether he'll do well in the near term. In other words, should you ride a hot streak (as so many people like to advise), should you do the opposite because the cold player is "due" while the hot player is bound to cool off, or should you ignore recent performance altogether and simply pick the guy with the best skill set as evidenced by his medium and long-term history. I used to be firmly in the latter camp, advising not to try and "time the market," arguing that because hot and cold streaks vary in length, it was impossible to predict whether a player was at the tail end of his hot streak or somewhere in the middle. And not knowing where he was on his hot or cold streak and therefore how long it would or would not continue, you could never use that information predictively. It was only backward looking.

But I'm coming around to the idea - or at least open to it - that maybe you should bet on the streak. Kevin Payne, who plays a lot of daily games, and Michael Rathburn, who used to work for Daily Joust, but is now at Fantasy Feud, both subscribe to factoring in a player's recent performance separate from his overall skill set. I argued with them, but both know what they're doing, and I found myself experimenting it as a criterion. Anecdotally, I have to say it seems to have worked - at least over 10-12 daily contests, no doubt an insignificant sample, but enough so that I think there's something to it. If I were to venture an explanation, it would be that you never know what's going on internally with a player, whether there's a latent, unreported injury, something in his personal life or maybe a mild illness through which he's playing but is costing him some focus and energy. So a player on a cold streak might be dealing with a handicap we don't know about. But a player on a hot streak is almost certainly not. At some point, he too could get the flu or have his wife cheat on him but the odds of that happening on any given day - when you know everything's going well for him the last few days - is slim.

Moreover, another theory I have is maybe players who are "locked in" are often doing well because their mechanics are sound. But over time, like most machines, there's some mechanical decay, and after a month or two, the player is in some bad habits and needs to go back and get a tune up, like when you get your car's steering realigned. Once they find themselves again they go on another hot streak until the decay sets in again. That means while a slumping player might be about to fix the issue and break out, it's impossible to see that coming in advance or know when that'll happen. But the hot player is likely to stay hot or only gradually cool down, so you can ride him until the breakdown in his mechanics are evident in his performance. Again, it's like your car - after realignment, you're good for quite a while. And once it goes bad, you might not take it into the shop right away - you'll probably just live with it until it's intolerable, and you have to fix it.

That's not to say luck isn't a big part in this, but you can see whether your player is flying out to the warning track and hitting line drives right at people, or whether he's completely lost.

Subject: Re: Charging
From: Andrew Martinez
Date: August 8, 2013 10:42 AM PDT
To: Christopher Liss

I think this might be my least favorite part of the baseball season because my mind is trying to stay focused on the task at hand (winning my leagues or salvaging respectability), but I'm being pulled in so many other directions as the PGA playoffs are about to start (some of us play fantasy golf), I have NFL fantasy football drafts to prepare for with keepers due, and college football is about to start, which means I need to prepare by reading as much information as I can about every team, so I feel like I have a good grasp of the upcoming season. This is when I really start to feel the dog days of Summer as a fantasy baseball owner, and I hate it because I love baseball, I just need more hours in the day to watch it.

On the topic of betting on hot/cold streaks vs betting on a player's skill set, when it comes to trying to predict future performance, I'm not sure where I come down. Let me explain...

On the one hand I see "streaks" as being a term we use to describe performances of which we have not yet figured out the cause. We say Nate McLouth is on a "hot streak" because he's getting hits more often at the plate than we are accustomed to seeing from him, but then his "streak" never really ends, and we come to realize he has raised the skill level at which he makes contact with the ball. On the other hand, Vernon Wells starts getting hits more often at the plate than we are accustomed to seeing from him, and we say he's on a "hot streak," but then he stops hitting and we're left to wonder why he got all those hits that he doesn't normally get. In the former case it really wasn't a "streak" and in the latter it really was, but at the time they were both labeled "streaks." The best fantasy owners are the ones that are able to separate the two and then use that knowledge in the market place to their advantage.

As for player skill sets, I have a difficult time always trusting them because they can change/evolve, and a player's level of performance within that skill set changes as well and not always in a linear fashion. Coming into this season I thought I knew how Kris Medlen would perform in 2013, given how he did in 2012, but then again I assumed he would bring with him the same skill set. Turns out I was wrong about how he would perform, not because of the results, but because of the process he used to get there as this season he's using a cutter, which is new, and now he's performing like a pitcher with a different arsenal. I don't have the luxury of a medium or long-term history, when deciding if I should start him or bench this version of him. The same can often be said of relievers who have skill sets and levels of performance that fluctuate, sometimes wildly. I thought I knew Fernando Rodney's skill set and had a good sized sample from which to judge it, until last season when I realized I didn't know his current skill set and what he was capable of achieving with it at all. The best fantasy owners are the ones that are constantly monitoring and evaluating player's skill sets and adjusting their level of confidence in that player to maintain that level of skill.

I don't play in daily leagues or formats, so I don't how much of this holds up in application. I would venture to guess the best owners in these games are thinking about all of these concepts on the fly, even if they don't realize they're doing it, and applying them to all of the options in the player pool.

Changing gears a little bit here, have you thought about what you would do, if you were Nelson Cruz and/or Jhonny Peralta regarding their PED suspensions? I find it a little bit odd that more is not being made of the fact that they chose not to appeal their sentences. I know I certainly would've tried to appeal because I would want to continue to play and help my teammates make the playoffs. The Tigers are fighting for the best record in the AL, and the Rangers are fighting just to make the playoffs, though they've now caught the A's, so it's not like their teams don't need them. I understand they're in the last year of their contracts and don't want this hanging over their heads going into free agency, but how many chances do you get to play on a team capable of making the playoffs? Next season these guys will be on different teams, in different divisions most likely, and who knows how good or bad those teams might be and the next time they might have a chance to make the playoffs. Also how dumb do the Rangers look in all of this for not planning for the possibility that Cruz might sit? The Tigers at least went out and made a trade that helps them through this transition and actually makes them better defensively going into next season. The Rangers knew this could happen and did nothing. I'm not saying they had to go out and get a major bat at the deadline, but to sit there and do nothing when your offense is struggling and you have plenty of resources from which to help yourself shows a lack of foresight on the part of their management.

Subject: Re: Charging
From: Christopher Liss
Date: August 8, 2013 10:42:40 PM PDT
To: Andrew Martinez

I think that's a good distinction between streaks which are seemingly random and new skill levels which are not. But that's a different point in that you're deciding what's sustainable and what isn't. But I'm talking about the ebbs and flows in a season for players of every skill level. Even if you're a .200 hitter, you're going to hit .300 for a couple weeks, and even if you're a .350 hitter, you're going to hit .220 for a stretch.

The question is whether you're more likely to go 2-for-4 after a .350 stretch then you are after a .220 stretch. Maybe the answer depends in part on someone's underlying skill set, but even so, I'd like to understand what kind of skill sets are streakier and what kind are more even-keeled.

As for the suspended players, it's easy to say that. But the difference could be millions of dollars, and we don't know their individual financial situations. I agree one should typically do the right thing out of principle, but it gets murkier when you consider how corrupt Bud Selig (baseball's commissioner who turned a blind eye during the Sosa/McGwire home run chase that helped resurrect baseball from a strike on his watch) is, and how sleazy some of the owners are (the Marlins fleecing Miami for stadium money, or Frank McCourt turning the team into an ATM). At what point does one conclude baseball is a business, getting caught for steroids are a cost of doing business and why should you and your family take a voluntary financial hit because Selig wants to clean up a mess of his own making? Plus with the Wild Card, five out of 15 teams make the playoffs from each league every year, so it's not all that rare.

Subject: Re: Charging
From: Andrew Martinez
Date: August 9, 2013 3:13 AM PDT
To: Christopher Liss

I don't know how I forgot that I had read this when you initially posed the question, but Tom Tango, Mitchel Lichtman, and Andrew Dolphin answered this exact question of streaks and their effect on future performance in their book, The Book. They did a study where they looked at 5-game and 7-game hot and cold streaks, based around, wOBA, and how batters performed in the 1-game and 3-games immediately following a streak and came to the conclusion that, yes, indeed players did perform slightly better after a hot streak and slightly worse after a cold streak, but that the difference was so small, it was basically meaningless (it was around five points of wOBA for hitters, after a 5-game streak, which I consider meaningless). They also did this study for pitchers and found that for starters, in the 1-game after a 4-game hot streak their wERA (wOBA converted into an ERA) was almost 0.3 runs lower than their expected wERA, which I consider significant. And after a 4-game cold streak their wERA was around 0.22 runs higher than their expected wERA. They also looked at the 4-games after a 4-game hot or cold streak and found that starters maintained some of their "streak" but to a lesser degree. They also did this for relief pitchers, but you can read up on than on your own time. So if we are to believe their study, and I find no reason not to, it turns out streaks have little to almost no significant predictive value for hitters, but streaks for starting pitchers do, at least for one game, after a streak of four hot/cold games.

I also recall a debate about this topic where Keith Law made the point that "locked-in is a myth" and that the "bunching" of hits is "just normal randomness" to which starting pitcher Brandon McCarthy took exception saying, "when science finishes completely understanding the human brain, then I'll defer solely to it." McCarthy's point relates to what you were discussing earlier in that you never know what's going on internally with a player. McCarthy went on to make some points about these internal feelings and how they can't be measured, at which point Tom Tango, one of our authors from earlier, jumped in and said McCarthy was "asking for it" and then tried to "teach him science" as to how it relates to being "locked-in" or "in the zone" and things devolved from there. You can read a better recap than what I just gave at the A's SBNation site.

While I accept the data, I find some merit to what McCarthy claims about internal feelings. We sometimes have this debate when we discuss closers and whether certain relief pitchers have the "stuff" or the "intestinal fortitude" to get the final three outs of a game. I think sometimes just because you can't quantify something doesn't mean it does not exist and there is something to being able to close a game out. We have science telling us otherwise, but science isn't on the field internalizing all of this. Then again, maybe it's better science can't measure those feelings and the objective view is the correct one as athletes don't fully understand what forces of probability and skill are at play here. I find McCarthy to be both fully understanding of the science and a believer in those feelings, so I find his comments on the subject more meaningful, than say your average athlete.

On the suspended players topic, if I'm a steroid user, getting caught is the price of doing business, but that doesn't mean I'm not going to hire the best lawyer I can to fight my charges and see if we can't get my sentence knocked down or my case thrown out. It's the same for a professional criminal, who gets picked up by the cops, and hires a lawyer to fight their case for them. And tell fans from KC, Pittsburgh, Toronto, or Seattle about how it's not all that rare to make the playoffs, I'm sure they'll have a different take.