2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

[Joey]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

Thanks, TT, I'll check it out!

Bahnzo's complete suite ...

http://www.operationsports.com/forum...available.html

Here's the spreadsheets you're looking for.

Again, much credit to Bahnzo, Knight, RCarew, and all of the others I am missing. Great work that makes a very good game, a great game. Shame on me for wasting 3 years trying to get 2k7-9 to "work right".

[Rwings832]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

If you see anyting that is off, please let me know. By that, I mean anything that is going to skew gameplay and results to a degree that it ruins it.

For example, early on I was using Runs Created for Clutch, but the results would have been Pujols with a cluth of 99, etc. So, my thinking was that anytime guys were on base, he'd be automatic runs, and that would ruin the gameplay. So, I scaled it to be between 55-75 for the most part, for both batters and pitchers.

What is your thought of relievers having K9 ratings in the 90s? Many of them will just because their K9 stats are above league average. I thought of having a sepearte scale/formula for relievers, but then that might have the opposite effect where they don;t strike out as many as they should, and you'd be better off just leaving in a tired starter, which we know is not good management.

---------------------------------------

To the other guys ... anythng I post should be considered "donation to the community". Use it however, whenever, you want. Improve it, alter it, whatever you want to do.

The more people making rosters, current, classic or otherwise just strengthens the community and the enjoyment of the game.

The way I understand the player vault for MLB11 is that once someone makes "Mookie Wilson", anyone can import him into their rosters. Then we'd just need to change his ratings for whatever year we were making. Theoretically, there could be a roster for every year of baseball, and it wouldn;t be impossible to do so. That would be awesome.

It is however customary to give credit to whatever roster you're using.So, for example, if one were tweaking Knight's rosters, they had better give him credit for the base work (or even ask for permission to use ... even though I forgot to). I posted Bahnzo's formulas under the premise that he relesed those to the public in his spreadshet.

I also posted them in such a way that you could just copy them into the formula bar in excel and double click on the stat abbreviation (or highlight it) and then click on the cell you want to use it on and it would change it in the formula bar. Then just copy that cell and paste it into the whole column and change the ratings for thousands of players in mere seconds.

The more poeple making rosters, the better. Lots of people have contributed to the process over the years, starting with Robert Glass and Paul W (and Swinging Soriano, P-Diddy, KG, Totte, and others) from the MVP and MVP edit days.

I'd love to hear RCarew's full thoughts on ratings and such, but I understand why he is reserved in revealing his methods.

Having an established and well-known consensus on the scales and formulas would be critical to all future roster endeavors. Quality control, so to speak.

Again great work TT, also good find on the spreadsheets, I don't have excel right now, so I can look at them, but trying to edit anything I cannot do. I agree with you I am hoping once players start being created, classic or current, popping out rosters and years should be pretty easy I would think.

[gtm]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

Most of us have used Bahnzos tool when it was working. So do any of feel that you will be making any changes to his formulas for more accuracy and then repost the formula lines you change?

[Rwings832]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

Most of us have used Bahnzos tool when it was working. So do any of feel that you will be making any changes to his formulas for more accuracy and then repost the formula lines you change?

Not sure if it needs to be drastically changed, probably will just need someone to remake the spreadsheets, or database so it can be used again from what I gather. If one can be made this would really help us out for the classic rosters.

[gtm]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

Not sure if it needs to be drastically changed, probably will just need someone to remake the spreadsheets, or database so it can be used again from what I gather. If one can be made this would really help us out for the classic rosters.

We really need to petition some of ratings guys to make this happen.

[Rwings832]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

We really need to petition some of ratings guys to make this happen.

yeah we do, definately have to put a bug in there ear

[TripleThreat1973]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

What you need are people that can work with C+ basic or whatever programming code was used for the program, specifically to pull the data into the spreadsheet.

All of the ratings guys can "make the spreadsheet", the formulas, and things of that nature. My guess is that a program like Access would be even better than Excel. What needs to be made is the stats retrieval and compartmentalization code that pulls the stats from baseball-reference, sorts them into the spreadsheet, and allows the formulas to covert them into ratings.

I'll talk with the Computer instructor here at our school, and see if he can help me do this for excel or access. Maybe we can at least get the spreadsheet working ... and working to where it can pull the data (via macros) for anything 2010 and before.

Good news is that a snow day has already been announced for tomorrow. So the 24 teams in the 2011 ZiPS roster will be completed by the end of tomorrow, then I'll just be waiting for the remaining 6 teams projections ot be released.

[Rwings832]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

What you need are people that can work with C+ basic or whatever programming code was used for the program, specifically to pull the data into the spreadsheet.

All of the ratings guys can "make the spreadsheet", the formulas, and things of that nature. My guess is that a program like Access would be even better than Excel. What needs to be made is the stats retrieval and compartmentalization code that pulls the stats from baseball-reference, sorts them into the spreadsheet, and allows the formulas to covert them into ratings.

I'll talk with the Computer instructor here at our school, and see if he can help me do this for excel or access. Maybe we can at least get the spreadsheet working ... and working to where it can pull the data (via macros) for anything 2010 and before.

Good news is that a snow day has already been announced for tomorrow. So the 24 teams in the 2011 ZiPS roster will be completed by the end of tomorrow, then I'll just be waiting for the remaining 6 teams projections ot be released.

Nice...yeah I am pretty tech savvy but am not a programmer and definately don't know C+. Definatley talk with you instructor, all we need really is the macros to work again and I think we will be golden.

[emath]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

I am not trying to say this is 100%, but I know someone who has done this type of thing in the past (rating creator/editor), and their thought is that maybe the guys at B-R.com aren't happy with people pulling the info directly from their site like that, and that may be why they tweak their site in order to keep programs like Bohnzo's from working.

I understand it makes your progress quicker, but it may be better to have a spreadsheet that just does the calculations, and you input the stats yourself.

[Nez477]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

TT-

The MLE model is so so important and will help you create an awesome set by the look of things, great job. I don't have '10 (only '09) so I won't be able to enjoy but this looks to be really really solid.

[TripleThreat1973]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

TT-

The MLE model is so so important and will help you create an awesome set by the look of things, great job. I don't have '10 (only '09) so I won't be able to enjoy but this looks to be really really solid.

I think you're right.

Just for kicks I did a fantasy draft using my 2/3 of the way done rosters.

LAA's Mike Trout was the 12th player in terms of "value". In terms of "bar length" he's the LAA 4th outfielder, but given his age (19) and cost (30K), his "value" is unreal. You won;t be able to trade for Trout without giving up a lot.

Meanwhile, if you decide Hunter and/or Wells are not worth their contracts, you could trade them and use Trout. Trout and Bourjos are going to cover a lot of ground. I'm thinking of doing that very thing, although I usually play the season with real trades, injuries, etc.

Jason Heyward is also highly valuable given his talent and age.

[TripleThreat1973]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

I am not trying to say this is 100%, but I know someone who has done this type of thing in the past (rating creator/editor), and their thought is that maybe the guys at B-R.com aren't happy with people pulling the info directly from their site like that, and that may be why they tweak their site in order to keep programs like Bohnzo's from working.

I understand it makes your progress quicker, but it may be better to have a spreadsheet that just does the calculations, and you input the stats yourself.

Baseball-Reference would probably love it if people used their stats for rosters and the like. IN a sense, they are competing with other sites that offer stats (Fangraghs, etc).

What they probably don't like are people pulling their stats through spiders or programs, and then using those stats toward actions (Projections, gambling, etc) that are profitable to someone else ... or another stats website basically using bb-ref's stats at their own site. In other words, letting BB-ref collect and manage all the stats, meanwhile you're website just pulls them over to your site.

[emath]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

Not the most web knowledgeable person, so maybe I am wrong, but would that affect their bandwidth at all? The person I spoke to said that there would probably be an issue for them (B-R.com) losing money.

It's possible he also meant them losing money on ads, since by automatically pulling the info via the editor, as opposed to people actually going to their site, and getting the info in person, they would be losing out on that income. Also, if no one is going to the site directly, they are most likely not donating money to the site... And yes, I do donate, as I had Joey Votto's page as mine (I sponsored it) until I had a money issue, and had to let it lapse before last year.

[Jaffa]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

TT: Could you post the formulas for movement or break of the pitches. Thanks ahead of time.

[gtm]

Re: 2011 Projection-Based Rosters: Methods, Formulas, & Discussion ...

Is anyone reworking Bahnzos formulas, if not would someone post all the formulas he used for pitching and batting. This will be going to excel so I need the correct way to enter the formula. I will then need to know what numbers I will need to input to get the correct aatribute numbers. Since we can't access Baseball-Reference with his roster tool.