Clay Davenport: Cespedes = Adam Jones

[Lucky Jim]

I'm not sure what you're getting at. Cespedes or a Japanese player aren't the same as someone who'd been floating around the minors for years and then suddenly tore up AA like Luis Montanez. They're players who were forced to play at a lower level because of local laws or customs.

If someone were transported from space to the BaySox at age 26 and started hitting .350 with power I'd certainly take notice.

[Stotle]

I have a lot of thoughts bouncing around in my head, and I've started to type this post three or four different times and kind of given up.

I guess it boils down to you know the error bars will be bigger, but are they so big as to make the data worthless? I tend to think not. I think there's a relatively static spread of talent in Japan/Cuba. Different players will be able to take advantage of the differing talent levels differently, but I'm not sure that's a game breaker.

You often read that a AAAA guy is taking advantage of crappy AAA pitchers, but gets used by real major leaguers, and I've always thought that was a little overstated. I don't think you have too many guys who hit .500 against AAA pitchers with 83 mph fastballs, but .177 against better pitchers, while others have a .300/.270 split. But that's just gut, little or no data.

It would be interesting to see a comparison from Clay on standard deviations of his translations between MLB-MLB, Japan-MLB, Cuba-MLB, and AAA-MLB.

Some of this is just common sense, no? Using a real world example, when teams try and utilize college stats to help with projecting out a player, particular attention is paid to the cross-sections of those stats that are most relevant to projecting a skillset, and those cross-sections are largely dependent on opposition type.

If Clay is going through and noting splits, including performance against pitchers that eventually became MLers, or pitchers throwing past a certain velocity, etc., then maybe I can buy in to there being some value to the projection. But taking stat totals and utilizing methodology to assign value as relates to others in the league, and then projecting out to a more difficult league, seems crazy to me. There is too much variance in talent for the stats to matter much.

Watch one Yu Darvish start and you can see how little should actually be drawn from his stat line if you are trying to evaluate his performance. His competition is all over the board.

[DrungoHazewood]

I guess we have different ideas about where the burden of proof resides.

I guess I don't see the harm in adding these translations to the dataset we know about Cespedes and using it as one input into a decision making process.

[CA-ORIOLE]

What are your data points? There have been, what, fewer than 20 pitchers that have come over from Japan -- all of varying ages and skillsets (reliever/starter, pitch type, etc.) over the last 15 years or so? What's the pool of Cuban players that have come over and performed in the US such that we can even start to determine what the actual standard deviation is?

I don't want to get into the statistical argument and probably lean towards agreeing with you here, but is there not enough scouting/observational information available to project him? Wouldn't this be a case where your expertise would theoretically provide an advantage over the statistics?

[Sports Guy]

I guess I don't see the harm in adding these translations to the dataset we know about Cespedes and using it as one input into a decision making process.

There is no harm in it...Treating it like a reliable piece of info is where the harm is.

[Stotle]

I think there are about 40 Cubans who've played in the majors over the past 10 years, and I'd assume a larger number who've played in the minors. 50-some Japanese players. The error bars will be higher than a AAA-to-MLB translation. But how much? I'm not willing to say they're so large that the information is useless.

You have a sea of useful data points for MiLers. Hundreds of players a year with careers that span over many years and levels.

The performance of 40 players (pitchers and position players) of varying ages and skill levels over a period of 10 years are going to give you anything useful in terms of determining how performance in one league might translate to performance in another? No way. If I gave you a list of 40 Major Leaguers over the last 10 years and their HiA stats and asked you to look at the current pool of HiA players and project their future MLB performance do you really think your projections would be in any way useful?

Sorry, this doesn't pass the smell test.

[Stotle]

I don't want to get into the statistical argument and probably lean towards agreeing with you here, but is there not enough scouting/observational information available to project him? Wouldn't this be a case where your expertise would theoretically provide an advantage over the statistics?

I personally think scouting carries much more value in the Cespedes discussion than an attempt to project future performance based off of his performance in Cuba. I'd go a step further and say the stats from Cuba alone are largely irrelevant unless you have someone familiar with the league that is able to put them into context. Not just a comparison against the other players in Cuba, but an explanation of how the game is played, what the general skill level is, what the general approach is, etc.

[DrungoHazewood]

Watch one Yu Darvish start and you can see how little should actually be drawn from his stat line if you are trying to evaluate his performance. His competition is all over the board.

But every pitcher in Japan's competition is across the board. Colby Lewis' competition was all over the place in Japan. Koji's competition was all over the place. But they were all excellent players who presumably had good translations. I guess you're saying that there are a large number of players who have good translations, and many/most aren't going to be any good because the skills don't translate. I'm not convinced that's the case. It's not like the Japanese leagues are the NCAA, where half the pitchers can't break 80 on the radar gun and any given team has zero MLB-quality players.

[DrungoHazewood]

I personally think scouting carries much more value in the Cespedes discussion than an attempt to project future performance based off of his performance in Cuba.

Agreed.

I'd go a step further and say the stats from Cuba alone are largely irrelevant unless you have someone familiar with the league that is able to put them into context. Not just a comparison against the other players in Cuba, but an explanation of how the game is played, what the general skill level is, what the general approach is, etc.

Not as convinced.

[Stotle]

I guess I don't see the harm in adding these translations to the dataset we know about Cespedes and using it as one input into a decision making process.

I guess. To me it's people taking a hammer and displaying how it's also useful as a tool for straining pasta.

[Stotle]

Not as convinced.

Open the newspaper in May and look at the high school baseball stat leaders in your county. Say the top 20 kids or so are listed -- cream of the crop. Now tell me what that means in the context of assigning future MLB value to these players.

[CA-ORIOLE]

I personally think scouting carries much more value in the Cespedes discussion than an attempt to project future performance based off of his performance in Cuba. I'd go a step further and say the stats from Cuba alone are largely irrelevant unless you have someone familiar with the league that is able to put them into context. Not just a comparison against the other players in Cuba, but an explanation of how the game is played, what the general skill level is, what the general approach is, etc.

So, out of curiosity to what type of accuracy do you think he can be projected based on the available scouting information that teams now have? What's your personal/professional opinion on his projection?

[LookinUp]

I think the value of stats for players like Darvish and Cespedes is in comparison to everyone else in the league. If your K, BB, ERA, BA, HR, BB totals are consistently at or above the top of your league, and you have good physicality, that means you're probably a much better bet to make it than, say, Connor Narron, whose senior year stats blew.

Now, how useful are those stats when drawing statistical comparisons to guys like Adam Jones, or when valuing a player's worth? That's very difficult to say, and that's where scouting and team-led workouts come into play.

In the end, I'm making the right Jones trade regardless of whether Cespedes is his statistical replacement. You don't trade Jones under the assumption that Cespedes will fully replace him, or frankly under the assumption that you can ever sign Cespedes.

[Stotle]

But every pitcher in Japan's competition is across the board. Colby Lewis' competition was all over the place in Japan. Koji's competition was all over the place. But they were all excellent players who presumably had good translations. I guess you're saying that there are a large number of players who have good translations, and many/most aren't going to be any good because the skills don't translate. I'm not convinced that's the case. It's not like the Japanese leagues are the NCAA, where half the pitchers can't break 80 on the radar gun and any given team has zero MLB-quality players.

Being "good" is a different issue than saying "Japanese player A will be most like MLB player B".

I think we can all agree that Yu Darvish will be "good" in some form. I'm not close to convinced that when Szymborski posted his ZIPs calculations for Darvish that they reflected a good estimate of Darvish's future performance.

[Stotle]

So, out of curiosity to what type of accuracy do you think he can be projected based on the available scouting information that teams now have? What's your personal/professional opinion on his projection?

I am not sure how you measure degrees of accuracy. I think someone who is familiar with Cespedes as an individual, familiar with the differences in Cuban/American culture on and off the field, and is also a good baseball evaluator, could fairly accurately describe Cespedes' upside and correctly identify the biggest hurdles he'll have to reach that upside.