Back to Boards

oR and dR


oR and dR

Are these Offensive Rating and Defensive Rating? I just saw these added this afternoon.

Re: oR and dR

it is our wonderful webmaster Guy going to work in his lab to shut talking heads up like me the only way he knows how. Its why we love him and are loyal to the site. #phantomoftheballpark

Re: Re: oR and dR

This looks like a great option.

I could of also suggest either to apply ranks based on position only.

Re: oR and dR

I will really have to borrow that hashtag!

TBH - I always planned this, but I probably didn't expect such an immediate need for it. You guys know I'm not one to dictate - if I push back it's because of one of two reasons: making sure a change works for a wide group of people and not just one small group that dislikes something, and of course my time - some ideas I just can't get to.

Re: Re: oR and dR

So far, I am absolutely loving this change and having this information readily available.

Re: oR and dR

Sorry to be the party pooper... but it is wrong and misleading and need to be fix...

Here is a perfect example in best of 2000 league

Jack Wilson : DWAR 4.07 fielding % .982 = Drank 92
Marco Scutaro: DWAR 1,96 fielding % .984 = Drank 89
Omar Vizquel: Dwar 0,75 fielding % .995 = Drank 90


Unfortunately... Jack Wilson is not in the same league as the 2 others defensiveley and should be rank 110 if possible... not 92

At 4,07 DWAR he is by far the best defensive SS in the 2000's league....
Vizquel despite is great field pct is worthless with is 0,75 DWAR...

I can careless about these rank numbers... I never use them before...

It should be there to help new players... but it doesn't help anybody if it is misleading...

As of now the new changes are totally misleading even more that it was before...


https://www.pennantchase.com/lgTeam..
https://www.pennantchase.com/lgFree..

Re: Re: oR and dR

plp, I don't think you understand Rank.

Re: oR and dR

It has been mentioned that dwar has larger ranges thus the way rank works is confusing if you don’t understand it. We don’t see this problem with other stats because there are not gaps in stats between two players like there are with dwar.

I can put some thought into how to resolve it but the stack rank process is what causes this.

I think the focus on new users has probably gotten out of proportion here. New users are still going to have to be willing to learn things about the game.

That said we’ll see if it leads to problems such as someone trying to claim a 4.00 dwar player is really only “2 rank points” better than a 1.8 dwar player.

Re: oR and dR

I am just reporting what is obviously wrong... so hopefully it can be fix...

Wilson is basically twice as good defnsively than Scutaro...
and 4 times better defensively than Vizquel...

but by rank numbers they are almost equal... and this is quite misleading...

if Rank is there to help new players and it is just misleading... well it does'nt do any favor to anybody... and it doesn't do it's job.

I am seeing tons of non sense in those numbers popping out.. and surely can't report every single one of them...

At least, I raised the flag about it being wrong and misleading... so it can be improved...

I am not criticizing here... I am just reporting some numbers to help out making more sense to these... so it becomes a relevant tool...

As of now... for me, it is irrelevant right now...

I also think it will have an effect on quick drafts.. and this is my main concern...

for me rank should not even be there... maybe it is too much time and effort spend on this...

it is on my part..

so I can't even imagine how much time it his for you Guy... I understand the complexity of the task...

Anyhow..

My own white flag is raised! I give up on arguing about it. Have fun with your rank system!

Cheers

Re: Re: oR and dR

Rank is an objective number based on some algorithm. Guy has indicated that both dWar and Fielding Pct. are being included. I do not believe position is factoring in yet though, which may be an area it could be improved more since a great SS/2B will bring more defensive value than a great 1B. It is also possible he still needs to fine tune the weighting of dWar vs Fielding Percentage.

Some assumptions I make about the Rank number based on what I observe in ATG auto leagues:

1. Rank is across all players, not by position. So the best SS/2B will not necessarily be ranked as 100

2. I assume Guy just orders all players from first to last based on his algorithm and if there were only 100 players the first would be rank 100 and the last would be rank 1. Of course there are a lot more than 100 players, so maybe the first 5 are ranked 100, the next 5 are ranked 99 etc.

It might not work exactly like this, maybe the rank "bands" are based more on the actual number he assigns so that some ranks get skipped entirely when there is a big drop off.

Anyway, when the objective rank does not match your subjective opinion it may just be a sign the algorithm still needs to be tweaked but there are always going to be outliers and it is not meant to be perfect.

Rank does not effect how the player performs in the sim.

Re: oR and dR

Basically, there's a sort now for what I used to do in my head 😎

- The Sheik

Re: oR and dR

Maybe this will help.

Most MLB stats fall into reasonably normal distributions. Thus, there is usually larger differences between numbers at the ends of the values than there are in the middle.

When you change this to a rank, the best number is always 1 and the next is always 2, no matter what the difference is. Particularly with dWAR, the difference between the best and next best can be larger than the difference between the say a 60 rank and a 50 rank.

So what rank will only tell you is who has the best value and who has the next best value etc. It doesn't tell you anything about how big that difference is.

There are other ways to do ratings but rank is simple and works pretty well as a simple system.

Hope this helps.

Re: Re: oR and dR

Velomonkey- you SHOULD be right however see the League REAL Stats Below and sort by DWAR

https://www.pennantchase.com/lgStat..
Carlos Gomez, Hughie Critz and Lou Boudreau are all a 100 Rank in Dwar however there are 43 points between Gomez and Critz and 145 Points between Gomez and Boudreau and they are both 100 Rank at Dwar

They should not be according to your logic, which by the way, I agree with

So I suspect that because Critz and Boudreau are Both SS and have a High Dwar the number of chances they will get at their particular Position is also figured into determining their Defensive value.

Oh and by the way, 3B get absolutely shafted in DWar, they don't get hardly enough robbed hits as they should an Adrian Beltre, Scott Rolen or Brooks Robinson should have a LOT more impact than they do in Dwar.

Look at Beltre in that League above a +2.31 Dwar and in 108 games has robbed 12 hits- that's criminal

Re: oR and dR

SCM,

The dR is not solely based on dWAR. It is a combination of dWAR FLD%. Unfortunately, catchers, first basemen and some outfielders will have close to 1.000, so the rank of great shortstops will be much lower, causing the drop in overall rating. Third basemen are the big losers in this system.

In reality, it is easy to come up with a metric that accurately comes up with an expected defensive output in PC, but I believe that Guy is limited to using published stats, so it has to be up the user to come up with that metric.

Re: oR and dR

It's missing the boat to continually harp on any form of Rank being "totally misleading" and the litany of other hyperbolic phrasing. Anyone who expects one number to reflect more than what it is intended to indicate doesn't understand that number's purpose.

Rank puts a pool of 200-plus players into numbers 1-100. Rank rolls together different stats into a single number. Pushing for Rank to separate players in a more granular way is a fundamental misunderstanding of how Rank is meant to be used.

Rank is not intended to do more than compare a pool of players in a general way. In general, Wilson is a better defensive player than Vizquel, and his defensive rank reflects that. Job done. Now, the human brain can analyze and evaluate and participate to decide how to use the rest of each players' stats in game play.

Allow the human brains to do some of the work, be it a new PC'er or a GM who has been here since the turn of the century.

Re: Re: oR and dR

Velomonkey, I think you missed my point, I really don't care what the DRank is - it's immaterial to how the player should preform and yes usually 3B have a lower Fielding percentage than SS and 2B

while 1B have close to a 1000 Fielding percentage a lot - it's easy to not make errors when you have zero range...

But a +2.31 DWar 3B SHOULD have a LOT more than 12 Robbed hits after 108 games and DRank has nothing to do with the Robbing hits formula, something is askew

Beltre has 282 chances ( 34 Put outs, 243 Assists and 6 errors) so his Fielding percentage is spot on

with a +2.31 Dwar he should have about a 9% chance of turning a hit into an out, but there is no way to tell how many HITS went to his direction to find out what his actual ROB rate is

Re: oR and dR

I think that's a sim question, not a rank question.

I agree with you that 3B defense is shortchanged in the sim.

Re: Re: oR and dR

All this makes me realize there are a lot of people here way smarter than I am.

Re: oR and dR

or and DR have not been added to the draft room i belive