Mastersball Fantasy Forum

Howdy,

Before I start a discussion on a strategy I may be becoming infatuated with, The Mayberry Method, let me ask first: Am I allowed to post on this? (Not sure about the rules for other competing and perhaps coined strategies).

Really dying to have a discussion on it. Especially since many on the board seem familiar with it.

-Will

for me personally, I think the MM is fair game. It's not as if Todd won't be a primary guest speaker at Shandler's First Pitch forums this spring. I haven't run any analysis on the method yet but I think [when push comes to shove] it will end up as a variation of tiering with certain players within each tier being primary targets.

The good thing is Ron is a friend and I can call or e-mail him and ask that he takes a look at the thread to make sure we are cool. Or one of his "spies" will put a bug in his ear and he'll shoot me a call or a note.

That said, the best place for a discussion on the Mayberry method is on the Baseball HQ forums themselves. Of course, that assumes you are a member.

For those reading this and are asking WTF? -- Ron Shandler introduced the Mayberry Method in the 2010 Baseball Forecaster. It is a system that puts a grade on certain skills to give a quick snapshot view of the player. For those that have read the description of the method and are familiar with how we give dollar values with category values (HR$, SB$ etc.), you may have asked yourself, "can't I just sort by category dollar value to get a lot of that info already?"

Yes, you can .

Not all, but the primary reason we provide categorical values is for this very purpose. The MM might identify an available speed guy or power guy. You have always been able to do that here by sorting on HR$ or SB$.

In the name of full disclosure, I am not the best person to be discussing the MM, or some of the tools JP has in mind.

Doing what I do, I become rather intimate with the player pool. The tools JP has in mind and Ron's Mayberry method are designed for those that love fantasy baseball, but have other things going on as well, like jobs, and spouses, and looking for spouses and raising kids and trying to make kids and other hobbies. What you are supposed to glean from the MM score, I already know because due to the nature of what I do with projections and profiles.

One of the reasons JP is such a good fit for us is he has a better handle on this simplification and chunking of information than I do.

Am I hinting we have plans to include some ideas of our own with respect to "information at a glance?"

Yes, this forum is ideal for that discussion and yes, the MM is fair game.

The MM is a simple reaction to the fact that there is a secret to baseball projections. There is science to figuring out what skills a player has but there is little science to how often a player will play. I like to compare it to Quantum Mechanics and Einstein's Theory of Relativity. Both predict with great accuracy their areas of science, but when you combine them they produce infinities. You know they work together just fine at some level because the universe is not in an infinite loop (or at least we think it is not). For BBHQ where they have spent an enormous effort into the science of predicting skills, the fact that playing time comes down to sitting around a table and saying "I think he gets 600 PA this year" is troubling.

It really comes down to playing time. If you buy a subscription to the Platinum Plan, you are really buying into our ability to predict playing time. Yes, there will be exceptions. Joe Mauer doubled his home runs. There is nothing in the numbers that said he would do that. However, using the tools of scouting told you that he was very strong and capable of hitting for great power. His power was hidden from the numbers.

So, if you are likely to be wrong guessing playing time, what can do you? You can ignore it and simply say he has the power to hit a home run every 20 PA. That is what MM does. The only problem is that you cannot assign a dollar value to MM. You can make a tier system like Viper said.

This does take the steam out of what we are going to be doing here at MB in some ways. We are coming out with our own coding system attached to the player profiles, but to provide a different type of information. I am working on an article on the Zen Projection Engine's apparatuses for determining playing time.

As part of the merger, we looked at the two projection systems, NEUTS and ZPE. We determined that the ZPE system for determining playing time was stronger, so that is the method we used for this year's projections.

IF we know that playing time is just an educated guess, then MM is a valid way of skirting the issue. I always liked the Baseball Perspective of saying someone will hit between .240 and .280 and if they do they claim credit for accuracy. Marcel says if a player hits well he should get more at-bats so ignore the drama.

We here at MB are old school. We will put our money on the table and say 630 PA. But, we will also tell you where we disagree (Jacoby Ellsbury) where we don't know (Jose Reyes) and where we refuse to speculate without facts (Stephen Strasburg).

A few points on this:

-My big positives on MM are: 1.) It can be adapted to other projection systems (I was thinking this the moment I saw the "up to x" piece).
2.) It looks at skills not stats. I know this is present in many strategies. But I don't yet see anything in the MM based on "Get this guy because he's projected for 40 and you can get him for 30."
3.) They use a broad range of skill for each tool with the "tier numbers".

What I would like to do is create a strategy that really throws away projected $. I don't think overall $ value is the best measure of a players value for the purposes of winning a league. This is obvious I'm sure.

I'm not saying it doesn't aim to simplify things and it probably takes out some intuition/skill from the fantasy game. But what if someone were to do a simple translation of mayberry values of skill with their favorite projection system and instead of translating that into dollar values of HR/SB/RBI/R/BA, he or she just bid as much as it took (within reason. No one should get suckered into buying a Carlos Gonzalez for 20some dollars... Maybe one could compare the average production of those with high MM #'s and set their max bid their.)

Anyways, just my take. I'm sure this will be paired with some $ value thing anyways. That'd be a shame though. And I know I could post this on the HQBoards but it's more low-key here and I like having access to the experts as well as a different perspective.

I'm not so sure I agree it is more low-keyed here but we all have our opinions. I got hooked on Mastersball when I got involved in a long discussion on DIPS. That thread made me a life-long member. The thread was long and detailed. Todd and others seemed to really know and understand the concept and were able to explain its inner-workings in a style that you could understand. BBHQ and Mastersball are the only two sites I subscribe to. 90% of my posts are here although I read BBHQ. I am waiting for the posting season to begin. It is probably more quiet here than there right now.

By low key I don't mean less posts or less activity. I am looking forward to more posts. More-so just the contemplative and thoughtful nature of the this board. I could explain further but they are just very different. Granted I've only seen the mastersball board a tiny bit in the last two years. Moreso in the last month.

I, too, only subscribe to BBHQ and MB. I've been a BBHQ subscriber for over 10 years and have seen BBHQ morph from being aimed at hard core roto gamers to now being aimed for the masses. As such, the BBHQ boards have a mix of very informed posters to novice posters, which, IMHO, can get extremly frustrating. I find the MB boards to be what BBHQ used to be and what JP refers to as "old school" which happens to appeal to me personally.

Just my 2 cents. Looking forward to the discussion.

Andrew

I too just read the MM piece in the latest forecaster. I thought it was interesting. Every year I tell myself I am only going to barnes and Noble to read it, but I always buy it. Don't know why as 3/4 of the book is the same every year..But I think I like reading the blurbs on each player..

The basics of what Shandler said is something I have come to realize in the last 2 years. We many times try to over complicate the process of assembling a winning team. You do need to properly value the player pool for your league type..Thats a Fact. But to use those dollar values as absolute bidding values is where the problem gets messy. And this is what I think Ron is trying to address here... the difference between these two projections below could easily be more than just a couple dollars depending on the league set up.

A.300 32 HR 105 RBI 15 Sb projection
B .325 38 HR 130RBI and 19 SB projection

Both Are legit Studs! Both represent roughly the same ability and skill set most likely...But in any given year, one may get more AB's, one may get more RBI chances due to batting order or the team he is on ...One could just be luckier..

But the idea is on draft day, the difference in dollar valuation could be as much as 5 bucks possibly. For essentially the same player.. One was luckier one year.. And this is exactly what I think MM is designed to do...represent a players tools and ability and group like players together. This is done so one can identify the pockets of value within groups.

Something I did last year was ditch all the draft software and BS and I came to the drafts with pen/paper and list of players in rank order by position...All run on CVRC and tweaked... I noted just a few things on the sheets.

1.Top dollar value for each position. 2.the Negative value line. 3 a few brackets within each position....... Thats about it and it worked amazingly well... I forgot about precise dollar values and just grouped like value players together...I also had some highlights on players I thought would represent value in each tier, but you get the idea..

For instance, on many draft lists I had for different leagues last year, I noticed S S Choo and Nick Markakis were relatively close to each other..I knew I could own Choo for half of what Markakis would cost...And I did in most leagues...And It turned out well...

So I think the MM way if used properly has some value. The value is in grouping players together and finding exploitable pockets of value on draft day...but you can do a similar thing by using decent projections and proper valuations and bracketing prior to draft day i think..

I am wondering if one could device an actual valuation system based on a simple system like this. Not to precisely assign dollar values, but to help identify potential profits within positions... I think someone once devised one according to ADPs from mock Draft central.... It basically took your available pool of positive players and dollars available and somehow devised a formula for translating ADP into dollar values...using an MM skills score ranking per position would be similar i think...

Anyhow, just some thoughts...

I've read the MM article a couple of times now. I think it's biggest flaw is that it is too granular. To show you what I mean, let's use Miguel Cabrera as an example as Ron does in the final article. Their first set of projections are up. They have Cabrera hitting 33 HR in 646 PA.

While in spring training, Cabrera breaks his leg during sliding drills. We know that a broken leg takes four to six weeks to heal. We learn that there are no complications and we have an expected return of the third week of April. We can reasonably estimate that he will miss 70 PA on the DL. Using the old fashioned method, we do the math (33/646*576=29). Cabrera looses 4 HR and is now expected to hit 29.

Using the MM method Cabrera didn't change in value. Cabrera's PX score didn't change. Cabrera will still have more than 500 AB so his MMPlayingTime didn't change. In reality, his HR, R, and RBI changed and if you are drafting, he fell a couple of rounds.

The other problem is that it is still completely dependent upon prior statistics. This method isn't going to help you find the next breakout like Joe Mauer.

That being said, I'm all for coding specific pieces of information that help add color to a set of projections. Health, Versatility, Defensive Ability are things that can be coded so that when you are bidding you can decide to spend that extra buck.