A bit of an epiphany regarding ratings/scouts in text sims.

[Ben E Lou]

I've mentioned elsewhere that while in most text sims I want to play more as a GM than a coach, spending lots of time getting the players in place, and spending some time, but not as much, setting strategies/game plans/lineups/etc. However, something always gets in the way of that. Being that text sims are, at their essence, glorified optimization problems, the things that keep my team's performance from being optimized are the things to which I end up paying attention. The epiphany I've had this morning is this: the thing that slows me down more than anything else is evaluating players based on their on-field performance.

I've been playing a lot of FBCB lately, and I'm using 1-100 ratings. Doing so is allowing me to be able to move from season to season rather rapidly, because I don't have to spend time after games determining if Player A or Player B is better for my system. There is no scout error, and the numbers are precise. The result? I'm getting through tons of seasons! I don't have to spend time after a few games deciding: "Hmmmmm, is Joe Smith's 5 rating in 3-point shooting is better than Frank Jones' 5 rating in 3-point shooting. Let me check their performances. Hmmmm. Jones has a higher shooting percentage. I'll start him the next five games and see what happens." (Five games later...) "Well, Jones shot 40.5% during those five games, but I'd better see what kind of defenses those teams were playing to be sure..." You get the idea... Instead, if I need an 3-point specialist, and Smith is rated 54, and Jones 50, I know who is better. I can move on quickly from there and look at other things.

I am fully aware that the thought of having 1-100 ratings with no scout error would be considered sacrilege to some. The good news is that this is the sort of customizable option that isn't of the type that I abhor: those that change the way the AI has to "think." (or those that end up limiting the game's depth for everyone....) This is the sort of option that could only have an impact on the way the gamer plays the game--not on how the AI must play the game. Heck, for a greater challenge for those who consider player evaluation to be an essential part of playing text sims, let the AI see 1-100 ratings, while the human could see 1-20, 1-10, or A-F.

I've realized that this is one of the little things I couldn't put my finger on that didn't allow me to get into TPF and OOTP as much as I'd like. In both games, in order to maximize your performance, it is an absolute must to spend time evaluating whether one 5 contact 6 power RF is better than the other, or which 6 pass rushing DE is better.

Bottom line: I think I'm going to enjoy OOTP6 a lot more with 1-100 ratings, and scout turned off.

[wig]

I've always thought that scout error was a stupid thing.

[Ben E Lou]

Originally Posted by wig

I've always thought that scout error was a stupid thing.

I think some form of variance/unpredictability is important for rookies and younger players, but for vets, it isn't very realistic.

That being said, if you like to have to evaluate players, I think the system in some games--where you "see" your players accurately, but see players on other teams with less precision--is pretty good.

[rexallllsc]

I think scout error is somewhat lame, but I also like to see variance in performance based on certain things. For example...is Kobe Bryant the best 3 point shooter in the league? Nah. Best slasher? Nah. Best rebounder? Nah. Best ballhandler? Nah. But something make him one of the best, if not the best player in the NBA.

It's nice to see guys with say, an 85 overall play awesome night in and night out, and then see another guy rated 85 play good, but nothing close to the first player...if that makes any sense.

To me, if all you're doing is plugging in numbers...it would get kinda old, yah?

[Anthony]

Originally Posted by rexallllsc

I think scout error is somewhat lame, but I also like to see variance in performance based on certain things. For example...is Kobe Bryant the best 3 point shooter in the league? Nah. Best slasher? Nah. Best rebounder? Nah. Best ballhandler? Nah. But something make him one of the best, if not the best player in the NBA.

It's nice to see guys with say, an 85 overall play awesome night in and night out, and then see another guy rated 85 play good, but nothing close to the first player...if that makes any sense.

To me, if all you're doing is plugging in numbers...it would get kinda old, yah?

i agree with rex all the way. at some point it stops being a game and becomes "let me see how many players i can get rated as close to 100 as possible." i like to simulate the issues and obstacles reall ife GM's come across to see if i could be as successful. yet - do real life GM's have infallible scouts who are never wrong about their evaluations on players? no, they don't.

[Ben E Lou]

Originally Posted by Hell Atlantic

"let me see how many players i can get rated as close to 100 as possible."

That's an over-simplification of where I'm coming from, though.

[Ben E Lou]

Originally Posted by rexallllsc

I think scout error is somewhat lame, but I also like to see variance in performance based on certain things.

Most definitely. I wouldn't want a 73 contact hitter to bat within 10 points either side of .290 every single season, for example. That's where proper randomization of the engine comes into play--which is another discussion entirely.

[Maple Leafs]

Scouting is probably the toughest nut to crack in sports sims. I've never seen one yet that really seems to do it well.

One thing I don't like is the process of bidding on scouts. In virtually every scenario, the human player will always outbid the AI on scouts because we know that good scouting is crucial to success. That's why I never play with scouts on in OOTP -- I'll always wind up with the best scouts, which is a huge advantage. FOF is a little better because at least the high-priced guys will do some damage to your bottom line, but we all still grit our teeth and pay them what they want, don't we?

(Not to mention that FOF uses several categories, as opposed to OOTP's ridiculous "hitting/pitching" breakdown. Is that being changed in v6?)

For a long time I've been arguing that we shouldn't be paying for scouts. Instead, we should be choosing what kind of scouts we want. Give every team a scouting staff and let them assign "points" to various areas in such a way that you're going to be strong in some and weak in others. Veterans vs. rookies, pro leagues vs. minors, currents vs. potentials, etc. You could be good in some areas, but not all. Now scouting becomes a game of strategy, not finances. The game that's come the closest to this model was EHM, by the way.

Another idea I'd love to see some day: Have player ratings/talents changing in the background before we see the numbers move. In OOTP, if my CF prospect improves his power talent, I know about it that day. I'd rather not. When I know should depend on my scout. If he's good, he'll figure it out quickly. If he's bad, he'll take a while. I think that's more like real life. If Pedro Martinez has three bad outings to start the season, the good scouts will figure out whether it's a slump or whether he'd just done. The bad scouts won't.

Think about it... let's say you're in rebuilding mode. You tailor your scouting operations to focus on young players and draft prospects. One day you get a message from your scout: "New York has a kid named Smith who's not well-regarded, but I see something in him. He's not projected to start for them, but if we can pick him up cheap he could be ready to really blossom." So you check New York's roster and sure enough the kid has so-so ratings. You offer them a decent veteran and they're more than happy to take him. Two months later, the kid's ratings have shot up. But in reality, they were already up... it's just that your scout was the first to "see" past the consensus. Think Priest Holmes, or Markus Naslund.

Same scenario, but in reverse. Los Angeles has a stud player that you've had your eye on for years. One off-season they come to you with a trade offer. Must be a salary dump, but you snap it up -- the guy's a superstar. Then he shows up in camp, and he's lost a step. Turns out his ratings had dropped last year, but the LA scout saw it before yours did. Your scouting staff just screwed up, and cost you big time. Think Ken Griffey, or Herschel Walker.

Now that would be a game where I felt like my scouts mattered. Right now, I just pay them as much as they want, and sit back knowing I'm seeing better ratings than the AI teams.

[MizzouRah]

Bottom line: I think I'm going to enjoy OOTP6 a lot more with 1-100 ratings, and scout turned off.

I've started my first ootp dynasty with coaches and scouts turned off, and yes I'm enjoying it much more now than ever before.


Todd

[sabotai]

Some people like to play "I'm going to sim this season and see what happens." Others like to play "I'm going to sim this game and see what happens."

It's good to see FBCB let's you do both.

[wig]

I agree with Maple. Scout bias is much better than scout error.

For unpredictability, I'd rather see it in player development instead of scout error.

[Ben E Lou]

Yeah...I LOVE Maple's idea about scouting points. The whole resource-management side of text sims would then come more into play.

[Maple Leafs]

Wig/SkyDog, have you guys ever played EHM?

It's not quite what I described, but it's close. You assign your scouts to various areas: minor leagues, overseas, etc. You can call them back at any time and reassign. So for example, you may have two scouts watching your team, one watching your farm team, three or four scouting various countries overseas (it's hockey, after all), one watching the US college system, and one watching whoever your next opponent is. Towards the trade deadline, you can pull a few of those int'l scouts off their assignments and have them watch various teams you're looking to deal with for a few games.

Not a perfect implementation, but pretty good. Too bad EHM was just too slow to really get into. Maybe the new edition that SI Games is working on will be the holy grail.

[cthomer5000]

Originally Posted by SkyDog

I think some form of variance/unpredictability is important for rookies and younger players, but for vets, it isn't very realistic.

Really? So why do so many free agent signings become complete busts?

[Ben E Lou]

Originally Posted by cthomer5000

Really? So why do so many free agent signings become complete busts?

I doubt that the reason is because a scout didn't know that they sucked. I'd be willing to bet that on most occasions, it is something more subtle like an offseason injury, getting complacent because of making the big bucks, not working out during the offseason like they should, off-field problems at home because of either moving the family, or being 1,500 miles away from the family, etc. etc. etc.

[cthomer5000]

Originally Posted by SkyDog

I doubt that the reason is because a scout didn't know that they sucked. I'd be willing to bet that on most occasions, it is something more subtle like an offseason injury, getting complacent because of making the big bucks, not working out during the offseason like they should, off-field problems at home because of either moving the family, or being 1,500 miles away from the family, etc. etc. etc.

So scouting error must account for a bare minimum of situations where veterans greatly underperform?

[Ben E Lou]

Originally Posted by cthomer5000

So scouting error must account for a bare minimum of situations where veterans greatly underperform?

I'm not sure it is a bare minimum, but I wouldn't think that it is the case the majority of times, either.

[sabotai]

Maple pretty much summed up my idea of how scouts should perform as well. Nicely worded.

[Honolulu Blue]

I guess I see scouts a little differently than the people who have chimed into this thread so far. In OOTP, the stats tell the story, and I only worry about the scout's evaluation when looking at rookies who have played litttle or none at the major league level, or as a tie breaker between free agents.

In FOF, when evaluating my own players, the stats tell the story. If I have a 37/37 RB that's averaging 4.5 yards a carry as a starter, then I assume he's doing his job and that I DO NOT have a crying need for a RB, despite what the recommendations screen may say. Again, I mostly use the ratings for players who haven't played very much.

That said, I don't like the 100-point system that FOF has. It's either too accurate or implies an accuracy that isn't there. I'd prefer almost any other system - grades, 1-10, 1-5, or even FOF Classic's bars.

The one thing I'd like to see in FOF is an ability to sort free agents by stats. I think this is closer to the way real NFL GMs see things - they assume the players with the best stats are the best, and use (perceived) ratings as tiebreakers.

[sterlingice]

Maple Leafs, that was an absolutely wonderful post!

SI

[Ben E Lou]

Originally Posted by Honolulu Blue

I guess I see scouts a little differently than the people who have chimed into this thread so far. In OOTP, the stats tell the story, and I only worry about the scout's evaluation when looking at rookies who have played litttle or none at the major league level, or as a tie breaker between free agents.

In FOF, when evaluating my own players, the stats tell the story. If I have a 37/37 RB that's averaging 4.5 yards a carry as a starter, then I assume he's doing his job and that I DO NOT have a crying need for a RB, despite what the recommendations screen may say. Again, I mostly use the ratings for players who haven't played very much.

That said, I don't like the 100-point system that FOF has. It's either too accurate or implies an accuracy that isn't there. I'd prefer almost any other system - grades, 1-10, 1-5, or even FOF Classic's bars.

The one thing I'd like to see in FOF is an ability to sort free agents by stats. I think this is closer to the way real NFL GMs see things - they assume the players with the best stats are the best, and use (perceived) ratings as tiebreakers.

Right.....

and that's player evaluation via stats...the whole thing I said I'd rather not do, due to time. I want to see my careers over decades and decades. That's my point exactly.

[JonInMiddleGA]

Originally Posted by Maple Leafs

Not a perfect implementation, but pretty good.

I'm curious though ... did the whole scout system in EHM not feel like a "lather-rinse-repeat" process for you after a short while?

Don't get me wrong, it was a different approach & I thought i was fine, but it didn't really strike me as all that exciting either. Even with my leaning toward preferring micro-management to macro-management, at some point it became rather tedious to me to periodically switch the scouts from one place to the next.

I had it down to pretty much a routine:
1) Hire as many scouts as your budget will tolerate well
2) Hire the best scouts you can afford.
3) Start the year by assigning your best balanced scout or a pair of good unbalanced scouts to grade your team. (i.e. 1 good offensive judge & 1 good defensive judge, not borderline insane "unbalanced" )
3b) Assign your next best scout to the FA pool
3c) Assign the rest of the staff to the various team around the league (I finally just started doing them alphabetically), reassign them to an unscouted team when their grades come in for their current location
4) Once the league is scouted, dispatch scouts to watch the draftable talent.
5) As scouting reports disappear with time, reassign scouts back around the NHL.

I don't know if I've explained that very well (it's been quite a while since I've played EHM) but it just kinda got ... well, tedious after a while. I mean, I knew exactly what I was going to do each season, it was just a matter of remembering to do it & then taking the time to do it. The best thing I can think to compare it to is setting/changing training options in CM.
It was kinda neat how it worked but after a while, it got to be (IMO) a drag.

I guess what I'm getting at is that I found the EHM scouting system to be different, but not particularly "innovative" either.

[MizzouRah]

I switched to 1-100 ratings in FBCB and well, I really like them better this way. I think I'll use the same in ootp6.

HR, thank you for letting us change that mid-dynasty. It's the little things, maybe simple, but stand out for me with FBCB.


Todd

[Honolulu Blue]

Originally Posted by SkyDog

that's player evaluation via stats...the whole thing I said I'd rather not do, due to time. I want to see my careers over decades and decades. That's my point exactly.

We'll have to agree to disagree on this because...

1) It doesn't take me very much longer to evaluate players by stats + ratings than it would merely by ratings,

2) I LIKE evaluating players by stats because I feel it's more realistic and immersive, and

3) I don't care about careers lasting for decades and decades. Roughly 20 years is about all I can stand unless there's some specific goal for me to shoot for.

[Karim]

Maple Leafs summed it up perfectly. There should be an allocation of limited resources to various areas of scouting. IIRC, TPF implemented this.

I can't really envision a sports sim without scouting. I never want the numbers to be absolute. Great ratings with poor performance or mediocre ratings with great performance - let me decide what's going on.

One thing I like more and more is Solecismic's graphical representation of ratings. It enables me to quickly see if someone's worth considering. All other sims I've tried simply have a list of numbers that I have to sort through by other means.

[Axxon]

Originally Posted by Karim

One thing I like more and more is Solecismic's graphical representation of ratings. It enables me to quickly see if someone's worth considering. All other sims I've tried simply have a list of numbers that I have to sort through by other means.

I agree with this 100%. In fact, the only times I even look at the numerical ratings is when I'm writing about a player. When the easter egg was announced, I'd never even realized that the numbers weren't there for draftees. I just never need exact numbers.

It's my favorite representation of any sports game I've played. One-nil had a close system but I hated theirs. It was like FOF but instead of a line it was a series of blocks and I felt compelled to count the blocks. I've felt no such need with FOF.

[Ben E Lou]

Originally Posted by Honolulu Blue

We'll have to agree to disagree on this because...

1) It doesn't take me very much longer to evaluate players by stats + ratings than it would merely by ratings,

2) I LIKE evaluating players by stats because I feel it's more realistic and immersive, and

3) I don't care about careers lasting for decades and decades. Roughly 20 years is about all I can stand unless there's some specific goal for me to shoot for.

Definitely agree to disagree because....

1) If I'm going to evaluate by stats, I am doing it constantly. For me, it means checking every non-superstar player after every game in football or basketball, or every 2-3 weeks in baseball, and comparing them to every single backup at their position's recent performance.

2) I'd agree that it is more realistic, but it isn't more immersive for me because...

3) It is much more interesting and immersive for me to see, for example, how I can do in a football career as I go through several different starting QB's careers, rather than 15-20 years, where I might have only one or two primary starters. In fact, it is at around the 20-year mark in football and baseball when I feel like I'm just getting the fun part started--when the all of the first generation of players have retired. I get the most enjoyment out of moving from one QB, finding his young replacement, deciding whether to suck it up and start the youngster as opposed to signing a journeyman veteran who'll probably play better at least initially, plugging the youngster in when the time is right, dealing with a few seasons' impact of a big-contract QB getting injured, recovering from that, etc. etc. etc. Plus, I like seeing many players' full careers--and the only way to do that is to cycle through at least three or four generations of players--maybe 35-50 seasons or so.


This is definitely a matter of personal preference, though. Perhaps number three above plays into why I enjoyed CAD so much, and now FBCB, despite not really being a college basketball fan. I enjoy seeing my roster completely turn over, and it happens a lot faster in FBCB and TCY.

[GrantDawg]

I think TPF has done about the best job, scouting wise, I've seen.

[Ben E Lou]

Originally Posted by GrantDawg

I think TPF has done about the best job, scouting wise, I've seen.

You're probably right with regard to realism (you see your own guys with more accuracy than you see other teams' guys). I just don't like the 1-10 ratings. I might have three guards rated 5 or 6 in run blocking, and I'm having to start a different one each week, check his stats, check the ratings of the guy he was blocking against, etc. etc. etc., in order to decide.

[GrantDawg]

Originally Posted by SkyDog

You're probably right with regard to realism (you see your own guys with more accuracy than you see other teams' guys). I just don't like the 1-10 ratings. I might have three guards rated 5 or 6 in run blocking, and I'm having to start a different one each week, check his stats, check the ratings of the guy he was blocking against, etc. etc. etc., in order to decide.

I'm not saying the rating style is right (actually not crazy about that at all). I'm saying how it handles scouting (money spent on different areas, with a cap set by the owner) is better. For ability display I really like CM, though it probably would be good for you. There is no overall average, you have to look at a huge number of factors to determine which player is right for your needs (then again, it is much more of a coaching simulator than you'd like).

[yabanci]

Originally Posted by SkyDog

Bottom line: I think I'm going to enjoy OOTP6 a lot more with 1-100 ratings, and scout turned off.

I'm just the opposite. I like scouting turned on and using the "talent only" option in OOTP where you don't even see the ratings. I wish you could even turn off the stars.

[The_herd]

As much as I hated the thought of using a stars system of rating players at first, especially for a football sim. The more I play EHM, the more I like the idea. I can easily tell how well, or poorly, a player is playing. I don't really use it for star players, I know they will play well more often than not, I use it more for the average players on my team. Its easy to be fooled by a player picking up a goal or a couple of assists, but when I look at the game ratings and see the guy that picked up 3 points only put up a 6 rating, it puts things in perspective.

Another thing I love about EHM is after scouting a player, on your team or another, it will give you a player or 2 on your team the player would play well with, if there are any.

[Axxon]

I'm playing FBCB with the 20 point system and I've always used the 100 point one. I tried this at someones suggestion and I'm really liking it.

I'm finding it way easier for me to play this way.

I'm not stressing on the actual numbers. I know some folks stress on which 10 ( I'm playing a crappy team ) is better but for me it's more stressful to sort the 51's from the 54's. I use the numbers more as an indicator than a fact and even if I know they are a fact ( ie ootp with scouts off ) I don't process them that way.

This is one thing that makes CM so cool. No one stat will tell you who is good at what. Most other games don't have this kind of situation. Everything is cut and dry.

In CM, you ask, is this guy a better dribbler than that guy? He has a 17 the other guy has a 16. He's better right? Not necessarily.

What is his balance? How about his pace? etc. So many ratings factor in that question to really analyze it would give you a headache. The game still gives you hard numbers but relates them so much that the numbers are only a part of the puzzle.

It makes me glance at the numbers, make some judgements and then play him and see how he compares to expectations. Then I judge him.

I have played this game forever. The style necessarily flows over to games that don't require this. It makes judging a lot quicker and while it may not be optimal, do any of these games require that???

The point I'm making is that barring an intricate skillset meshing ala the mature CM, the 20 pt system in FBCB is great. It's precise enough that you really only have a +/- 5% error and it lets you quickly know the score.

If you go with the meshing then precision isn't a bad thing. Just throw in some intangibles like they do and you're set. We need to be able to trust the numbers but we can't make the numbers rule the sim. There must be need of human judgement. Judgement not necessarily analysis. Without judgement, the game would become stale to me.

[dixieflatline]

Originally Posted by SkyDog

You're probably right with regard to realism (you see your own guys with more accuracy than you see other teams' guys). I just don't like the 1-10 ratings. I might have three guards rated 5 or 6 in run blocking, and I'm having to start a different one each week, check his stats, check the ratings of the guy he was blocking against, etc. etc. etc., in order to decide.

That is a real problem with the small scales like 1-10. I think though that maybe if I was given a simple scouting report as to how well my players played the last game(like CM) that less swapping out of players might be neccessary. Preseason is the time to to sort out which of the 6 rated run blockers is best. One guy wins the job and unless he falls flat on his face he probably will keep the job for the year. This seems to work somewhat more realistically I think. If you have a star player who is a 9 in run blocking he job is safe in preseason. The guys fighting for the roster spots are three that all have 6. Game scores would be very useful not only in figuring out which of the guys is best but also if he is playing to that level in the regular season. Stats can be used for this but I will admit I don't have the patience to be looking through all of these stats to determine things.

[spleen1015]

The problem with scouts, especially OOTP, is that every scout sees Sammy Sosa differently. In real life, we know that Sammy Sosa can potentially 45-60 HR every year. There's not a scout in the world who thinks otherwise. In OOTP, one scout can rate him as a 6 in HR, another an 8, another a 9.

[dixieflatline]

I'm bringing this thread back because I want to discuss one other important job of scouts that didn't really get talked about much in this thread. From ESPN about the bears signing Ruben Brown:

There are mixed opinions as to how well Brown is playing at this point of his career. Some scouts feel he still has a couple of productive seasons remaining. Others suggest that the former Bills star has been in decline for the last few campaigns.

How much a player has left in the tank or what are the chances that a player will fall off the cliff I think may be the most important part about scouting free agents. I'm going to borrow a line here but adding something like the players chance to breakout, stay the same, or collapse to a player card would be very useful.

As pointed out earlier in this thread judging a players established level of play isn't hard. The big questions that the scouts could be used to answer is how much does Jon Lynch, Duece Staley, or Ruben Brown have left. Steve McNair had several years of slightly above average play and then the light turned fully on and he became a star. He could have just as easily stayed at his previous level and been a decent QB. Or maybe collasped like Kordell Stewart. FOF seems to do a decent job of having some players boom/bust around year 5 of their careers but adding boom/stay the same/bust percentages to the player card would be very useful. And maybe in addition to a young talent rating for scouts FOF could add a vertern collapse rating as well.

[Maple Leafs]

Originally Posted by dixieflatline

How much a player has left in the tank or what are the chances that a player will fall off the cliff I think may be the most important part about scouting free agents. I'm going to borrow a line here but adding something like the players chance to breakout, stay the same, or collapse to a player card would be very useful.

Good points. I don't really agree with the idea that everyone knows what the established players can do, so scouts aren't that important.

Look at Sammy Sosa. He's been dominant for years, but had an off-year (for him) in 2003. Was that a fluke season, or the beginning of a decline? I think a great scout could tell you the answer. A bad scout will just wait and see, by which point it will be too late to do anything useful with the information.

[condors]

well after many hours playing sports text sims....

Puts on SAAP fanboy hat...

i like no ratings at all, i didn't agree with this at first but it grew on me
With the scouts in saap now comparing that player to current ones in your squad and giving you the overall thoughts on the type of player your looking at. I think it is currently the best game with regards player development.

...takes off fanboy hat

Basically let me know enough information from my scouts to work off of and i will go from there. I don't want to know if guy has a 49 or a 51 i want to pick the ryan leaf and fire my scouts in fustration.

it is definatly an aquired taste and i don't look down on a game for doing it with numbers though.

[druez]

I need coaching and scouts in a game for me to feel like its a real simmulation. I think all games should include the option to do both.