|
||||
|
|
OOTP 20 - General Discussions Everything about the newest version of Out of the Park Baseball - officially licensed by MLB.com and the MLBPA. |
|
Thread Tools |
01-22-2020, 01:09 PM | #1 |
Minors (Single A)
Join Date: Jul 2017
Location: Germany
Posts: 50
|
Scounting: Favor Tools vs. Favor Ability vs. Neutral vs OSA - Here are the numbers
Hello guys,
before the latest Hotfix, OSA was the way to go when it came to scouting. After the Hotfix, I was interested in getting a deeper insight to finally answer the question for myself "What kind of scout should I hire?" So I decided to do some basic calculations to compare different scout types. I took the 2019 draft class in my current save game as the sample. Lacking of time I reduced the sample to the top 40 Batters and top 18 Starting Pitchers. I know that is quite a small sample, when it comes to statistical power, but I think it might help to give a brief overview. I underlied different scouting preferences and different scouting budgets and compared it to 100% accuracy scouting setting by calculating standard deviations for each scouting type. I did my caluclations only for the overall potential ratings. I know that one should consider other ratings for deciding to draft a player or not. In my opinion the overall potential is a good "core measure" for a quick calculation and for simplifying things. Furthermore, it is important to note, that I used 20 to 80 (Increments of 5) rating scale and high scouting accuracy, as these are my prefered game settings. I know using 20 to 80 rating scale is not the most accurate rating scale and results might therefore be biased, but I wanted results which are related to my "everyday game settings". So here are my calculated standard deviations: [1= 5 Mio Scouting Buget for amateur scouting; 2= 10 Mio Scouting Budget for amateur scouting, all settings with 200 "Scouting Amateurs" scout rating] OSA: 8.69 (batters: 8.14; pitchers: 9.79) Neutral 1: 9.65 (batters: 9.39; pitchers: 10.21) Neutral 2: 8.88 (batters: 8,48; pitchers: 9.71) Highly Favor Tools 1: 7.13 (batters: 7.29; pitchers: 6.77) Highly Favor Tools 2: 6.57 (batters: 6.93; pitchers: 5.65) Highly Favor Ability 1: 10.79 (batters: 10.55; pitchers: 11.30) Highly Favor Ability 2: 11.16 (batters: 10.81; pitchers: 11.90) --------------------------------------------------------- The following settings are apart from the previous "perfect world" settings: Favor Tools, 5 Mio, 170 Scouting Amateurs: 8.51 (batters: 8.40; pitchers: 8.74) Favor Tools, 5 Mio, 200 Scouting Amateurs: 8.95 (batters: 9.29; pitchers: 8.16) Highly Favor Tools, 5 Mio, 150 Scouting Amateurs: 8.64 (batters: 9.08; pitchers: 7.55) Due to some restrictions I mentioned above I am cautious in drawing any implications. For me it seems that OSA is still quite okay, especially for low budget teams which cannot afford a high-end scout for their franchise. My results suggest, that a mixture of scout and OSA should be a good choice for making draft descisions. Only having a high-end scout which highly favors tools seems to be superior. Basically these results give only a brief insight but I hope that this might help some people. Last edited by LeftHandPath; 01-22-2020 at 02:35 PM. |
01-22-2020, 01:26 PM | #2 |
All Star Starter
Join Date: Apr 2019
Posts: 1,085
|
Interesting, thanks for running this experiment. It seems from your findings that OSA is still pretty powerful in predicting the overall potential of players. My anecdotal thoughts seem to confirm this, with the caveat that the OSA individual potential ratings (CON/GAP/POW, etc.) often seem kind of wacky.
Would be interesting to see the difference with a max scouting budget ($24 million I think). Looks like in any case that the Favor Ability scout was a bit weak, at least in your experiment. I would have thought that the Ability guy might have kept the standard deviation down, but maybe missed some guys at the ends of the bell curves. I like a Neutral guy myself, but maybe I'll check out a Tools guy sometime. Also wondering how important Tools/Neutral/Ability is compared to their actual scouting ability for Amateurs/Minors etc. I'm in 2 online leagues, with one of them using no ability/potential stars or ratings, and the other league using the full gamut of numbers between 20 to 80. I would say the no ratings league is my preferred way of playing against others... that way you really need to know what you're looking for. |
01-22-2020, 01:43 PM | #3 | |
Minors (Single A)
Join Date: Jul 2017
Location: Germany
Posts: 50
|
Quote:
The scouting budget (5 / 10 Mio) is the pure budget for Amateur Scouting. 10 Mio Amateur Scouting budget is 42% of a maxed out 24 Mio budget, which is quite high. I think I will also chech the standard deviations for a weaker scout with Highly Favors Tools, as my results suggest that the focus seem to be more important than the scout's rating. |
|
01-22-2020, 01:55 PM | #4 |
All Star Starter
Join Date: Apr 2019
Posts: 1,085
|
Ahh, OK.. you're right that the $10m pure amateur scouting budget is on the higher end.
Something I thought of later is that maybe it makes sense that the Tools guy is coming out on top in terms of standard deviation, because you're only looking at the top 58 players in the draft. Maybe the Ability guy will be more accurate for 3rd/4th rounders and later. But then that begs the question, if the players outside of the top 60 or so in the draft never make the MLB, is the accuracy of their scouting report really that important? One thing I do like about having a good scout versus OSA is that your scout will be quicker to spot Talent Change Randomness. I saw it in one league where a GM put a pitcher on waivers who had just received a healthy TCR boost. Ouch. |
01-22-2020, 02:11 PM | #5 | |
Minors (Single A)
Join Date: Jul 2017
Location: Germany
Posts: 50
|
Quote:
OSA is the only one with some overestimates. I just checked the sample with a Scout which highly favors tools, 5 Mio Budget and 150 rating (which equals "excellent"). The results tends to show that such a scout is at least as good as a high-end scout with "neutral". Last edited by LeftHandPath; 01-22-2020 at 02:14 PM. |
|
01-23-2020, 11:20 AM | #6 |
Minors (Single A)
Join Date: Oct 2019
Posts: 53
|
Thanks for doing this. This community is fantastic, but I think the next project we should all work on is draft analytics. There’s so much we don’t know but want to know. I’ve been meaning to run a test on late round draft picks. Specifically whether it’s better to draft prospects with balanced potential across the board or players who your scout will say be above average at two things but pathetic at the rest.
Essentially in the 15+ rounds, should you take a batter with 40 potential in each category. Or should you take the batter with 60 potential in two things but 25 or 30 in the rest? God I love this game |
01-23-2020, 11:25 AM | #7 |
Minors (Single A)
Join Date: Oct 2019
Posts: 53
|
Thanks for doing this. This community is fantastic, but I think the next project we should all work on is draft analytics. There’s so much we don’t know but want to know. I’ve been meaning to run a test on late round draft picks. Specifically whether it’s better to draft prospects with balanced potential across the board or players who your scout will say be above average at two things but pathetic at the rest.
Essentially in the 15+ rounds, should you take a batter with 40 potential in each category. Or should you take the batter with 60 potential in two things but 25 or 30 in the rest? God I love this game |
01-23-2020, 12:39 PM | #8 | |
All Star Starter
Join Date: Apr 2019
Posts: 1,085
|
Quote:
For pitchers I like a balance across the board as a minimum baseline. If they're overly weak in any of the 3 main rating categories they will probably suffer. Two exceptions: Control can be 40 or 45 if they've got really hot Stuff and Movement. And Stuff can be low if you're playing in an earlier era and have a good defense. But modern day baseball requires good Stuff from pitchers from what I've seen -- looking for 55+ for a good SP, and 65+ for a reliever. For batters in the later rounds, for me it depends on my park. If I have a high AVG park I'm just sorting by Contact and looking for guys that have Contact plus something else interesting -- IF/OF Range, Catcher Ability, Eye, etc. Maybe even Speed for fun. |
|
01-24-2020, 12:21 PM | #9 | |
Minors (Single A)
Join Date: Oct 2019
Posts: 53
|
Quote:
|
|
01-28-2020, 11:45 AM | #10 |
All Star Reserve
Join Date: Aug 2002
Posts: 580
|
Nice testing. I also play with High accuracy and the 20-80 scale in steps of 5 so I was doubly interested in your findings.
I always go with Heavily Favors Tools scouts. Seems that they would do better when looking at players who are all judged solely (or very heavily) on their future value. Based on your results, it seems disappointing and literally pointless in terms of the POT rating unless the scout is top tier. That is disappointing. His "flavor" for finding potential didn't even matter unless we're talking 200 rating, which, I assume, is going to be rare. Granted, that last sentence may apply to me since I always invest in top quality scout, so that's promising at least, but assuming I can always find a top-tier Tools scout. Otherwise, it doesn't seem to matter as long as I'd avoid Ability, which makes sense as the draft is going to be a ton of unrefined talents. I guess the other "hope" would be that Higher Favors Tools would be better at judging the individual potential ratings. If anyone had the time for that, I'd be highly interested, especially since after a while, everyone is looking fringy (or worse) so it's about picking out who could evolve with some breaks later on and what "strengths" they might have already. |
01-29-2020, 04:29 PM | #11 | |
Minors (Double A)
Join Date: Feb 2019
Posts: 100
Infractions: 0/1 (1)
|
Quote:
So thank you, you've helped me. |
|
01-29-2020, 04:36 PM | #12 | |
Minors (Double A)
Join Date: Feb 2019
Posts: 100
Infractions: 0/1 (1)
|
Quote:
But maybe not so afterall Last edited by ThePride87; 01-29-2020 at 04:37 PM. |
|
01-29-2020, 05:33 PM | #13 | |
All Star Reserve
Join Date: Aug 2002
Posts: 580
|
Quote:
That would make a lot of sense, too. But I think that would need a change in how players are generated - some more prospects that are closer to their potential but not with a ton of upside, or at least that's how an Ability scout might see them...good at seeing what they can do now and reasonably expect for improvement, but miss like you said, the Trouts or the guys who might "get it" suddenly, etc. You'd get "pleasant surprises", less busts, but give up on some "dark horses" or "late bloomers". Right now, it seems like Ability = Actual (which is why they are weaker in draft), Tools = Pots, and accuracy...it doesn't matter unless the scouts are at or near 200 and Tools. It's just all disappointing. And...good enough...the Tools scout can hit on actual ability just as well. IDK it just seems weird. Accuracy might need to be more nuanced and less reliant on budget? |
|
01-30-2020, 11:23 AM | #14 |
All Star Starter
Join Date: Nov 2009
Posts: 1,703
|
I had been using neutral or favor ability most of the time. However, I switched to 'Favor tools' for my last two tries and have found a few differences.
My Favor Tools guy suggests more CFers and SSs than the Neutral guys. Balanced guys come up with more decent hitting corner players. The Tools guys really give me more Speed and Defense suggestions, so SS and CF makes sense, since they are two of the five tools.A lot of strong arms in this group. I still consider CONTACT potential to be the most important Tool for every batter. If you can't hit, you ain't getting out of the minors. I generally sort by CONTACT potential, then look for PATIENCE. With Tools scouts, fast guys with good defense don't bring up as many patient kids as before.
__________________
"My name will live forever" - Anonymous |
01-30-2020, 12:24 PM | #15 |
All Star Reserve
Join Date: Aug 2002
Posts: 580
|
Interesting. I use Heavy Favor Tools always and I get a mix of positions and Speed/Baserunning in his suggestions. In fact, some guys I outright reject as they have terrible defense, especially range. I value range on defense like you value contact for hitting.
If anything, I get a lot of pitchers suggested. I've had shortages of SS and CF in some years and would only get pitchers and 3B suggested a lot. The 3B are often the ones with terrible range and often the CF I think are more like corner OF eventually (40-50 range) Contact usually isn't a problem....it's mostly that he gives me Contact + AvK and little power. Eye hasn't been a strength of my team over the seasons either, so that's one similarity, but then I don't draft for it often when I override either so that's not all his fault. Or he'll get hooked on those 60 Power and nothing else type players in later rounds. I have high defense on the team at SS, but I also select for it myself, so I don't know if it's any impact of the scout's preference. My scout would likely suggest a SS that can hit but has 55 defense skills and I'm like "no, I want someone with 70 range at least, that guy you want would go to 1B or 2B." Last edited by KBLover; 01-30-2020 at 12:26 PM. |
01-30-2020, 05:41 PM | #16 |
All Star Starter
Join Date: Apr 2019
Posts: 1,085
|
I think the biggest test of an OOTP player is in online leagues. A good chunk of the GMs are going by the same conventional wisdom that you are, so you have to somehow figure out some sort of edge. In the single-player mode you can kind of just steamroll the AI if you know what you're doing. Except in the playoffs... (I'm salty)
Most statistical analysis done on OOTP has said that Contact is the best stat and that's likely true. You kind of have to have at least 55+ CON in most cases to add value with the bat (there's some exceptions like Gallo). But man, I think Eye is really important. It really adds a lot of value. I think it's more consistent too. Even if you've got a good contact hitter you can still get screwed by BABIP in any given season. I believe Eye is a more consistent rating that will see an obvious and expected increase to players' OBPs over the course of a season. I realize this isn't really anything that we don't already know. Sabermetrics figured this out a long time ago. But it's kind of funny that some people still overlook Eye just like teams did throughout most of baseball history. To tie my rambling post together, Eye is something I'm focusing on in one or two of the online leagues I'm in to see if I can get an edge as a small market team. Moneyball back to its roots. |
01-30-2020, 07:03 PM | #17 | |
Minors (Double A)
Join Date: Feb 2019
Posts: 100
Infractions: 0/1 (1)
|
Quote:
|
|
01-31-2020, 11:13 AM | #18 |
All Star Starter
Join Date: Nov 2009
Posts: 1,703
|
I do get a mix of suggestions when I press the SUGGEST button, it is simply that the suggested players often are IMPOSSIBLE to sign, so I ignore them. Then I do my sorting and usually find SS and CF look best to me.
I'm terrible at finding starting pitching, so I'm taking suggested starters in the double digit rounds. With a NEUTRAL scout, I sorted by EYE and then look for CONTACT. Since switching to a TOOLS scout, I've not seen many high CONTACT guys with high EYE as I'm used to. That is, high EYE rarely accompanies high CONTACT. Might simply be the randomness of about 10 seasons of play, so I'll see.
__________________
"My name will live forever" - Anonymous |
01-31-2020, 03:02 PM | #19 |
Minors (Double A)
Join Date: Feb 2019
Posts: 100
Infractions: 0/1 (1)
|
Just thinking out loud, but with the info that was discovered here, it appears OSA is still pretty solid on its own. There's still an edge to be gained when you have a top-tier tools scout, but with major league scouting, I wonder if I'm better off allocating $0 towards ML scouting and just rely on OSA....since ML-caliber players are already sort of "figured-out", I'd expect there to be a lot less deviation between OSA and the true ratings. But an 18 yr old? I'd rather shove even more of my $24M scouting budget into the amateur/minor scouting to figure out what their true potential is...we know potential for a youngster can be VERY volatile, so why not focus more on where the foggiest info lies?
This topic has really helped open my eyes a bit about scouting and questioning what I used to believe. I need to be careful, I know this was a minor experiment and I'd be wise to do my own research (OOTP is tremendous for figuring things out on your own, credit where credit is due) Last edited by ThePride87; 01-31-2020 at 03:04 PM. |
01-31-2020, 03:14 PM | #20 |
Hall Of Famer
Join Date: Apr 2015
Posts: 7,167
|
all that does is put a facade over scouting error -- which is based upon your budget and scout. as long as you aren't below baseline, you can safely assume your budget and scout are a high % correct than the osa.
the fact that OSA can be right about some is irrelevant. If your setup has a higher % chance of sucecssfully rating a player, it will always be a higher percentage, and without pulling back the curtain you will never know when it isn't at that particular moment of decision. so, the osa being right, but at a lower % doesn't mean anything useful other than to avoid looking at the OSA, which will only add confusion. so, if a good budget and scout is a higher likelihood of being accurate, you should never, never use the OSA... that's like going to vegas and betting against the odds. you will lose more often doing that than betting on the higher percentage chance. as far as the favor this or favor that... all that does is put a fake facade above and beyond the scouting inaccuracy based on the prospects age or whatever else they use to differentiate between tool and ability. ---- the way eye works in ootp... not the same as real life. or not quite. everythign is part of the same whoel and has to add back upto 1. the better the player is, the lower you want his eye rating --- up to a point. ... let's say a lower eye reduced bb by 10 (keeping it simple, this is actually a rate, not a #). those are now PAs that have to be divied up into outs, singles, doubles, triples, home runs, and anythign else i am missing in proper ratios as dictated by the ratings and competition faced. (errors are calculated/applied after a ball goes into play and needs to be considered, but only 1.5% chance on average, so it doesn't reduce those new opportunities much) as long as the resulting ball in play outcomes outweigh those lost 10 walks, it is beneficial to your team in ootp -- no argument possible if you know these facts about your player and league context. a walk is supposedly worth .33 runs (***your league may be different than RL this is also based on average of the league... if you have a better than average team, you should expect this value to be higher, since more opportuniteis and better percentages of following through with those opportunities compared to average with a 'better' team. that's a rising tide. a better team also increases the value of a HR, since people are on base above league average on your 'better' team) but, keeping it simpler, you only need the equivalent of 3 runs from 10 walks to get to the break-even point where less walks is better for that player. ~average team context why is this not quite like RL? well, in ootp you can get quite extreme about this with fewer negative repercussions. in RL there have been few players that can swing a bat like a wild, feral animal and have more than 1 or 2 good seasons before pitchers destroy their career with significnatly more sophisticated approach. vlad guerrero is one. and like the guess above, he's an elite talent that can get away with it, but likely would have been better with a bit more discipline, unlike in ootp results. same reason why jim furyk would have been a better pro golfer had he learned a more repeatable swing. impossible to say it isn't true. his swing is more prone to error all other things remaining the same. being wild at the plate makes you prone to more errors and mistakes - greater opportunities to screw up for no reason other than an unsophisticated approach. ootp is a giant pie. if you take 10 away over here they get shifted to the other possibilities in proper proportions. use this concept you can quantify what is happening and know for certain, with a bit of research/observation, as to what is best. it always has to add back up to "1" after an incremental change and all other factors remaining the same. you can predict with confidence. this helps use the LTM better to, if you fiddle with those. changing one is easy to predict what happens... but start moving more than one and yo h ave to understand where those 'lost' 100 hr go and how that extra '5000' walks league-wide reduces AB, which will reduce ball in play numbers... so that 100 is more likely 150+ lost hr if you make these 2 changes etc etc... |
Bookmarks |
|
|