Charging the Mound: At What Point Does a Hot Streak Become a Breakout?

-----Original Message-----
From: jeff@rotowire.com
Sent: Tuesday, April 10, 2012 4:55pm
To: "Christopher Liss"
Subject: Re: Charging

I had my last draft of the season last night, my original AL-only, 4x4 scoring home league auction. My first thought after the draft was "thank God my drafts are over." But whatever – it's a good life, and that's not what this column is going to be about. Instead, I want to talk about short-term fluctuations in value.

No matter how many times we preach not to get overly influenced by small samples, especially early season or spring training results, it happens all the time. Take the example of Francisco Liriano, whose value has been all over the map thanks to his results last year, in winter ball, in spring training and then in his first start against the Orioles this year. To give an idea of Liriano's fluctuating value, here's what he cost in four of my auction leagues:

• XFL – A 15-team, mixed, 5x5 keeper league populated mostly by industry types, with a couple of other veteran players mixed in. The auction is done at Ron Shandler's First Pitch Arizona Symposium during the Arizona Fall League, in November. He went unmentioned in the auction, but then was selected in the reserve draft in late March.

• AL LABR – 12-team AL-only, 5x5 auction league during the first weekend of March. Brad Evans took a chance on him for $6.

• AL Tout Wars – 12 team AL-only, 5x5 auction league, three weeks after LABR. Liriano went for $13.

• Amici – My aforementioned 11-team, AL-only, 4x4 home league from last night. He didn't go in the auction, instead falling to the second round of the reserves to Peter Schoenke.

Now, the first and last leagues aren't perfect comparisons, but they're close enough for this discussion. Good pitchers are good pitchers, and even though the 4x4 league doesn't reward strikeouts, I guarantee you that if Liriano didn't bomb in his first outing Saturday, he would have gone in the auction. I point out the purchase prices not to criticize those that spent more or less, but to illustrate that even in leagues populated by experts (or, if you prefer, members of the fantasy industry and fantasy veterans), we get swayed by short-term events.

Maybe Liriano isn't a perfect example because of his health issues last season, but then again, we always have underlying reasons when we change opinions on a player. My question here is this: Is it rational to react strongly to a short-term sample? Or are we just creating profit opportunities for others?

I think there are cases where it's better to act quickly. Maybe the Liriano owner in Tout Wars should be trying to trade him and still get $7 of value for his $13 player? If you were the Liriano owner, and someone offered an $8 Max Scherzer (the price is only relevant insomuch as it's how the market valued the two pitchers) to you, would you take him? It's easier to decide in shallower leagues with these guys on the margin – if someone demonstrates a potential high ceiling, you churn your roster to get him. But in deeper leagues, those guys aren't freely available, so you have to make harder decisions. But I also think in some ways you have to act more quickly. If Liriano has three more bad starts, he's going to be throw-in material only for his owners.

Do you agree with that premise, even if not necessarily with Liriano? If so, who else should we be looking to sell short while we still can?

-----Original Message-----
From: "Christopher Liss"
Sent: Tuesday, April 10, 2012 8:59pm
To: jeff@rotowire.com
Subject: Re: Charging

The 4x4 league is pretty deceptive because Liriano pitches for the Twins, who aren't projected to offer a lot of run support, and starters are mainly there for their wins. He might help in ERA and WHIP, but I'm assuming you count his first outing, so you're already in the hole there. I would have taken him in the auction for sure, but probably only for a couple bucks in that format. In AL LABR/Tout, I still think he's worth about $8. He was very good this spring, and we all know what he can do if healthy. In other words, I would have bumped him up from about $7 in early March (really should have bought him at the auction but had a bad vibe, so let Evans get him at $6), to $10 in late March (Tout), and back down to $8 now. In other words, I think his value has changed a little, but a couple good or bad outings aren't dispositive of much when you're dealing with a pitcher like that who could be great at any point, or get hurt and be out for the year.

I think we should be more prepared to act on a small sample where the magnitude of the change is greatest. In other words, if you flip a coin 10 times and get eight heads, I wouldn't conclude from that sample that the coin is weighted. But if you got 10 heads in 10 flips, the sample would be equally small, but the odds of that happening by pure luck are less than 1 in 1,000. It wouldn't be proof the coin were weighted, but it would at least make you seriously wonder.

Likewise, the size of the pitching or hitting sample can be offset by the magnitude of the change. If Johan Santana struck out 15 in his debut and averaged 92 mph with his fastball, that it was one game wouldn't matter so much. If Justin Morneau went 4-for-4 with a head-first-slide stolen base and clocked his head against someone's knee, but got up without incident and hit a homer in his next at-bat, we'd move the needle a lot.

In most cases, changes of that magnitude don't happen over a small sample, but there are exceptions. A closer winning the job is a change of major magnitude. A career minor-leaguer like Bryan LaHair homering one day and having two doubles the prior day has a bigger magnitude than anything Curtis Granderson or Carlos Gonzalez might do in a week.

One way to profit is when other owners don't properly adjust sample size for magnitude of change. Another is to more accurately gauge the magnitude of particular developments than other owners.

Slow starts from players like Colby Rasmus or Eric Thames whose job security is tenuous worry me more than a slow start from Kevin Youkilis who merely looks old. Youkilis – so long as he's healthy – isn't in danger of losing his job, so he'll have more of a chance to get into a groove. The sample of his poor play isn't large enough for me to conclude he's washed up. I could be mistaken about that, but that's how I assess the magnitude of their respective slow starts.
As for players I'd want to sell, there's a difference between those I could sell at their original prices and those I could sell at their current market prices. I'd be more inclined to buy Youkilis at what might be a discounted price than to sell Rasmus at a major discount. Of course, if someone wanted to pay Rasmus' original acquisition price, I'd probably get rid of him.

-----Original Message-----
From: jeff@rotowire.com
Sent: Wednesday, April 11, 2012 8:26pm
To: "Christopher Liss"
Subject: Re: Charging

You're right, the 4x4 league isn't a perfect comparison for the reasons you mention, and because of one nuance in our bidding structure. We're allowed to alternately buy our reserves, or do the opposite and spend all of our budget without filling the active roster, instead filling it during the reserve draft. The end result is that stars typically get priced a unit or two higher, as do closers (going hand-in-hand with the 4x4 format), and end-game guys instead get taken as reserve picks. The average team ends up having to fill two-to-three spots in the active lineup with reserve picks, though RotoWire founder Herb Ilk usually is in the opposite camp, having saved up enough money to not just fill his active roster but two of his four reserve spots in the auction. So Liriano's price is skewed a little by that.

I think that the changes in magnitude argument makes sense, though it's similar to the "signature significance" moments we've talked about in the past. But even though these outlier performances demonstrate a player having more ability than we once thought, his aptitude to regularly harness that ability might not come in the near future. I'm thinking of Luke Hochevar as an example, who I believe has indeed turned the corner after having a few of those outstanding performances two or three seasons ago.

But even though that aptitude might not come this season, the cautious route is not the way to go. If you wait-and-see if it's for real, you'll never get the player. I've made this mistake too many times in mixed league formats, and hopefully I've learned, after missing out on the likes of Jose Bautista, Ben Zobrist and Cliff Lee in their breakout years. We want to preach patience, caution and restraint, and that might still be right in many cases when it comes to cutting bait, it's a loser when it comes to targeting breakouts. So go grab the Jeff Samardzija's of the world while you still have the chance.

-----Original Message-----
From: "Christopher Liss"
Sent: Wednesday, April 11, 2012 11:29pm
To: jeff@rotowire.com
Subject: Re: Charging

Right – signature significance is part of that, but over one start the magnitude has to be incredibly high for it to be sufficient in and of itself (I'm talking 13-plus Ks, no walks against a top team). But even if the start falls short of that like Samardzija's first one, you can use the contextual clues (great arm, improved control in the second half, more experience) to give it more weight.

And of course we're dealing with uncertainty and small samples, so a lot of these guys on whom you gamble will come back to earth. But you have to keep putting your chips in when the potential payout is big enough to justify the cost, given the odds. The trick, of course, is getting those variables right.

I don't think it's an easy call whether in a 15-team mixed league you should drop a John Danks, for example, for Hochevar, Danny Duffy, Jake Arrieta or Samardzija. It's easy to say you should have targeted Bautista now, but Chris Shelton sure looked like he was headed for a breakout a few years ago, too.

Once you know enough to be patient with the players you've drafted and then realize that you can do that to a fault, you're left with looking at everything on a case-by-case basis. Being aggressive is foolish if you drop a reliable guy in a slump for a long shot lottery ticket. Being overly cautious and not grabbing players clearly showing improved skills is also dumb. I wish I had more actionable advice, but I'd say be cautious where it's warranted and take chances where it's not. There's no substitute for good judgment.