Baseball Reference and Baseball Prospectus, ugh.

[Colin Wyers]

The thing that aggravates me the most is that there is a single place you can go to see WARP3, Win Shares, OPS+/ERA+, and RCAA/RSAA. You go to BP for WARP3, THT for Win Shares, B-R for OPS+/ERA+ and CBE for RCAA. I suppose the reason is that each site fears loss of readers of their info is accessible elsewhere.

Yeah, but do you really need the whole alphabet soup?

For example, let's take Runs Created, Batting Runs and Equivelent Runs. RC I can get from THT or B-Ref, Batting Runs off B-Ref, and EQR (or BRAR/BRAA if you prefer it in those terms) from BP. How much of that do I actually need. Do I need RC/9 and OPS+?

At some point it's not more information, it's just more clutter.

If I was desiging a site from scratch, here's what I'd have (just discussing offense, leaving out pitching and fielding for the moment):

• All the plain vanilla event stuff that you have to have - hits, doubles, home runs, hit by pitch, GIDP, steals and caught stealings, etc.
• Your basic old-fashioned Runs and RBIs, just for nostalgia
• The triple slash - AVG, OBP, SLG
• Batted ball data - GB%, FB%, LD%, BABIP
• Runs. Use the best runs estimator you have available. Since you're using a computer to do all the heavy lifting, you should have some very good run estimators available.
• Convert those runs to wins.
• Give me ONE rate stat. And make it a good one - wOBA would be nice.
• Then give me runs and wins, adjusted by position. You can use average or replacement as your baseline, I don't care.
• Then a nice baserunning metric would be a nice way to round things out. Dan Fox has some great work in this field.

If you do decide to do replacement level, then I would like for there to be somewhere on the site where you explicitly spell out what the replacement level is for every season. Give me a translated player card for what a replacement level player might look like over a full season. Tell me how the replacement player compares to the average.

You could argue for a few more things to go on there - rate stats like K/BB, K/9, BB/9 for one. And if you have the data, something like Win Probability Added would be great to have too.

I don't need five measurements of the same thing when one will do. Quality is infinately superior to quantity.

[SABR Matt]

If I were designing a website, I'd have tabs for different info types.

First tab is basic data (including the triple slash stats ,OPS+ etc)
Next tab is That data converted to account for era.
Next tab is PBP level data like batted ball trajectories, ball/strike data, PBP baserunning analysis
Next tab is Basic Fielding
Next tab is PBP analyzed fielding
Next tab is uberstat analyses...all the major flavors
Next tab is split data

That way your view isn't cluttered on one big scary screen and people can search for the data they want, but all of the data is there.

[Tango Tiger]

If I were designing a website, I'd have tabs for different info types.

First tab is basic data (including the triple slash stats ,OPS+ etc)
Next tab is That data converted to account for era.
Next tab is PBP level data like batted ball trajectories, ball/strike data, PBP baserunning analysis
Next tab is Basic Fielding
Next tab is PBP analyzed fielding
Next tab is uberstat analyses...all the major flavors
Next tab is split data

That way your view isn't cluttered on one big scary screen and people can search for the data they want, but all of the data is there.

Exactly my point-of-view. Just like Amazon, Ebay, Paypal, etc.

[Tango Tiger]

Also with BBRef, I don't agree with the methodology in averaging ERA+ and OPS+ by calculating the career average on base, and slugging, and league ob% and slugging. This is a mathematical fallacy of averaging proportions. Its like saying that 5/4 (1.25) and 3/2 (1.50) average out to 4/3 (1.33). It brings good pitchers down over the long run and keeps good hitters higher over the long run.

That's not what Sean does.

Do you have evidence otherwise?

[brett]

That's not what Sean does.

Do you have evidence otherwise?

Well, yes, if you take their player ERA and league career ERA you get their ERA+ which is the exact problem I am discussing. If you take Walter Johnson's seasonal ERA+ scores weighted by innings, he comes out to be 161 which is more accurate.

With their methodology, all things being equal the lower run environment seasons get weighted less by their process.

[Patriot]

You can't weight ERA+ by innings, because the denominator is earned runs, not innings, due to the fact that it's league ERA divided by pitcher ERA. The proper way to figure career ERA+ would be to weight the individual seasons by park adjusted ER, or to find the league ERA weighted by innings and make the calculation at the career level.

[Tango Tiger]

Well, yes, if you take their player ERA and league career ERA you get their ERA+ which is the exact problem I am discussing. If you take Walter Johnson's seasonal ERA+ scores weighted by innings, he comes out to be 161 which is more accurate.

With their methodology, all things being equal the lower run environment seasons get weighted less by their process.

Tom to God: Please tell Sean Forman that most people are not aware that you cannot do anything with ERA+ unless they first take its recipricol. ERA+, unlike every other stat, puts the league figure in the numerator instead of the denominator. So, in order to do any kind of adding of ERA+, first do 1/ERA+, then weight it, then take the recipricol of that. Thank you God.

God to Sean: Make it so.

[Tango Tiger]

I created a wiki entry for this:

http://www.tangotiger.net/wiki/index.php?title=ERA%2B

[SABR Matt]

LOL

That was an amusing post Tom.

Is that what B-R does? Just out of curiosity.

[stejay]

I much prefer Baseball reference. i always have done. Never really got round to bp to be honest with you

[Tango Tiger]

LOL

That was an amusing post Tom.

Is that what B-R does? Just out of curiosity.

They do lgERA / playerERA, where lgERA is the ballpark adjusted ERA.

And when they do the career, they simply do the same thing: lgERA (weighted by the pitcher's innings) divided by careerERA.

The math is fine. Just the idea of putting the league number in the numerator is silly. It confuses at least 90% of the people when they try to do additions with it.

Even worse... he does OPS+ for pitchers as well, but he does it the right way (league value in the denominator). So, if the reasoning for doing ERA+ is for the bigger-is-better, why not for OPS+ as well?

And OPS+ and ERA+ track each other so well. A 70 OPS+ would yield a (real) 70 ERA+. Now, a 70 OPS+ would give you a 140 ERA+.

I vented to Sean on his blog, but I'm the only one.

[SABR Matt]

LOL...I can understand the frustration...but I did the same thing when I created DNRA+. I inverted it because it's easier to explain the scale to people that way.

Nonetheless, I see your point.

[Tango Tiger]

LOL...I can understand the frustration...but I did the same thing when I created DNRA+. I inverted it because it's easier to explain the scale to people that way.

Nonetheless, I see your point.

But, nobody seems to have a problem with ERA itself being a smaller number is better. That's what we expect from our pitchers. An ERA of 1.12 is a good thing, and an ERA of 7.67 is a bad thing. Why not simply say that Gibson had an ERA at 40% of the league average, rather than the league being 250% of Gibson?

[SABR Matt]

Fair point.

This is a case where Forman got wrapped up in his own idea of displaying things in an easy to see 100 is average scale and then so infected the population with that expectation that it has too much momentum to alter now.

Which is why I got infected.

And why people seem to understand a 150 ERA+ better than a 67 ERA Ratio.

[Tango Tiger]

The 100 infection is fine. I just don't get the 150 preference over 67 for ERA, but that it's OK to have an OPS+ for your star pitcher at 67!