[Syd Thrift]

I'm a big fan of stats only and this year is better than ever with the improved (and yet more concise) scouting reports. For those I generally use:

Actual ratings: None
Potential ratings; None
Other ratings: 2-8

Other ratings are the fielding ratings, including arm/range/error/turn DPs, which I think IRL are generally observed rather than looked up (I mean, so are power/contact/etc., but those are also generally only observed in game situations and as such mostly reflected in the stats), speed/baserunning, which also have a "you need to see it" component above and beyond what you can see in the boxscores or by playing out the games, and bunting.

In order to get the AI to behave like you, you also want to set the following in the AI evaluation ratings:

Ratings: 0
This Year: 67
Last Year: 22
2 Years Ago: 11

You'll still be able to peruse scouting reports and if there are no stats available whatsoever (for instance, when you're starting your league) the AI will default to ratings. Because of that, I still set Scouting Accuracy to Very Low. If you hire a really awesome scout they might be able to give you fairly accurate reports on most players but

The one downside is that your history reports will show Contact:: --(--) instead of Contact 55(60). Absent a way of generating 20-80 ratings purely through the stats themselves there's not really a way you can get around that.

The one time I do turn on scouting (potentials only though) is when I'm drafting because I don't use feeder leagues and as such the stats that do get generated are pretty random. Again, I trust the Very Low scouting rating to give me a lot of false positives and false negatives.

So... why do I and other people do this? We get an annual thread on this but I may as well proselytize here:

- In real life, you don't just get to look at a player's ratings and say "oh hey, this guy has fallen off the table, time to cut him". Sometimes players do just completely lose it over the course of one offseason in both real life and in OOTP, but it should be a really tough call to make if you have a 38 year old on your team who was still an average starter last year but who through April is hitting .150. Is he just in a bad slump? Or is he really and truly done?

- To that end, by the way, lest we think that actual coaches and managers have *that* much more insight than the stats say, consider this fact about football (different sport, I know, but the issue still applies): when you look at quarterback aging, like plot the average rise or fall in a QB's stats from when they're 22 to 23 and so on, roughly half of a QB's drop-off comes in their last season in the league. One way to look at that is, maybe guys play well and suddenly lose it, which definitely happens sometimes (Peyton Manning?). Another way to look at that is, maybe 36 year olds just have bad seasons the same as 26 year olds do, only when 36 year olds slump, they don't get another chance. I'm sure it's a combination of those factors but my point is, this is one of the dilemmas of managing/GMing that to me is the essence of what these kinds of games ought to be about.

- IMO potential ratings are fraught with issues too. Like, what we call "power" is actually a combination of, like, 50 things, from bat speed to pitch recognition to the average angle a player's bat has when moving into the strike zone and so on. Scouts try to do what they can to combine all of that into one number, but it's just plain not the same concept as what OOTP calls power potential, which is straight up "how many HRs this kid will hit in the future". Sometimes scouts can see that a big but skinny player will eventually become big and strong as he matures, but I'll accept that trade-off in favor of having no idea whether or not that 19 year old who's raking for me in the major leagues is the next Ken Griffey Jr. or the next Claudell Washington.

- I feel like, too, what Very Low accuracy does in the game is add a lot of variance to scouting reports. IMO what it *should* do is cause a scout to consistently underrate or overrate a player from report to report to report until it becomes blindingly obvious, perhaps from years of major league play, that that 15 HR a year guy does not secretly have 80 POW or that guy who strikes out a batter an inning actually does have a killer curveball. We're all only human and one thing that humans do the crap out of is we judge a person's entire life by our first impressions. What I see the game do (and correct me if I'm wrong on this) is that for players with little experience, the game is just plain off with the ratings. A guy with a "real" power of 50/80 might get a 60 in one report and a 45 in the next, with the spread slowly lowering as he gets more playing time. In an ideal system every single scout should have a little database associated with him that gives him a semi-permanent modifier. Perhaps some scouts might tend to overrate patience in general or underrate movement or whatever, but above and beyond that a scout ought to behave like a human does and that means putting a player into a category and only taking him out of that category when it becomes obvious.

- Also, too, while it's neat that scouting accuracy varies, you can know more or less how much it varies by the quality of the scout you get, and you basically know exactly how good the scout that you're hiring is. I mean, on the one hand I get that players aren't going to want to scour through a list of 100 potential hirees to review their histories or whatever but at the same time I don't think you should just get to be like "oh hey, Bob Jones is Legendary at evaluating rookies so I can more or less depend on him". IRL not very good coaches/manager/etc. get hired because of nepotism, because someone in the organization worked with them and likes them, and so on. Sometimes a guy who really hit on a couple of players was just lucky, or used to be really good but over time has developed some biases that make them not so great anymore. I don't think you should know all that stuff in advance. In an ideal system you might have to ask your assistant or your manager or whatever, all of whom are going to have different priorities and friends and so on.

***

Anyway nobody asked but all that's why I prefer stats only, at least for now.