Player Ratings

[Rumplebeanskin]

Please please please........No 99's.

It could perhaps be one of the biggest annoyance of mine in the whole of sports gaming. I might be alone, but I may as well put it on here. The existence of the 99 overall player.

Michael Jordan. Field Goal percentage of .497. He'd probably be a "99 Player", but he missed 50% of the shots in his career. Wayne Gretzky had a shooting percentage of 17.6%. He'd be a "99 Player", but he isn't perfect either.

I'm just trying to say, 99 is the ceiling that in my opinion, nobody should be reaching, because as good as anyone is, they aren't perfect athletes. Are the ratings and abilities of each player going to be out of 100? From what I've seen of the alpha build videos on youtube, players appear to have numbers between 1-100. Can we please have it so someone like Peyton Manning is at tops, a 90 overall? There is just no need, when you have numbers between 1-100, to squeeze everything to the top end of the scale, and make any number between 1-50 totally redundant.

[Francis Cole]

Generally you won't get any stats above 90, and above 80 indicates this player is supreme at the rating.
In the current version there are only 5 QBs with a Passing Accuracy of above 80 for example.




If you see a player at 99 this is basically saying he is perfect at whatever the rating is. I can't think of a single player who is flawless at a given skill.

[Rumplebeanskin]

Originally Posted by Francis Cole

Generally you won't get any stats above 90, and above 80 indicates this player is supreme at the rating.
In the current version there are only 5 QBs with a Passing Accuracy of above 80 for example.

If you see a player at 99 this is basically saying he is perfect at whatever the rating is. I can't think of a single player who is flawless at a given skill.

This has made me incredibly happy

[Rumplebeanskin]

And linked to player attributes, will player overalls be variable from team to team dependant on their style of play?

For example (very rough example indeed), DeMarcus Ware is an OLB Blitz Specialist. His stats will remain the same regardless of who he plays for. Would Ware have the same overall rating for the Cowboys (who would use him to blitz a lot, and covert blitzing linebackers) compared to say, the Bears who play a cover 2/tampa 2 style of defence and covert linebackers who can drop into coverage?

Same attributes, same ability, but the Cowboys would see more value (and as such a better overall level of skill) from Ware than the Bears would.

Granted this would require some sliders to be built into each coach and the importance they have traits at specific positions. WR catching vs speed (west coast vs air raid). QB arm strength vs arm accuracy. Kicker leg strength vs kick accuracy etc.

[Francis Cole]

I have no idea of how/if/when I will do overall ratings. Because of the exact ratings you say.

In Baseball it's easy to say this pitcher is rated 80, because general you can put any pitcher into another team and (aside from the DH rule and park factors) he will perform the same. But in Football how do you do that. A player playing in one scheme/team who produces the quality/results you would expect of an 80 player might be awful on another team. Without some serious thinking/work I can see it leading to nothing but annoyance to the user.

[Rumplebeanskin]

Agreed, and thanks for the quick response.

It is a can of worms, but ultimately a very important one.

I think one way to address it may be to have internal modifiers to certain attributes depending on the preference/scheme a team/coach is set as using. This could be pre-set, or it could be a sliding scale.

I am a coach who operates a power run game, and as such covert power backs. I 100% value strength over speed (or you can make it a say, 75% to 25% weighting, so it isn't absolute, I still need my back to run under 6 seconds for a 40!). The engine can then subtract a predetermined amount behind the scenes to stats like speed, agility, acceleration, and artificially inflate stats such as strength and trucking (if it's included). I still see the stats of a player as they really are in the user interface, but the number used by the engine to calculate a player's overall rating are significantly altered. If you have a slider to say how you you prefer power or speed, this could just alter the amount the internal modifier affects an attribute.

If I then plug said player into my roster, he should excel. His stats externally haven't changed, but I see a high overall because he is good at the things I need him to be good at. Shonne Green probably wouldn't be a better back than Chris Johnson, but he would be if heavily used a power run game plan. If I then put Shonn Green on the Eagles and their high tempo scheme, he would tank, even though his attributes haven't changed.

I think the tricky part is balancing everything.
How important are certain attributes to certain player types?
How much numerically should an attribute be booter/reduced?
What number would an attribute(s) need to be at roughly for one player to excel in a suited scheme, and under perform in a contradictory scheme?

Can of worms indeed!!

[Francis Cole]

Originally Posted by Rumplebeanskin

Agreed, and thanks for the quick response.
I am a coach who operates a power run game, and as such covert power backs. I 100% value strength over speed

If that was the case wouldn't you just look at a players's power attributes and not their speed attributes anyway rather than relying on an "overall rating"?

[TGH-Adfabre]

I would look at them with or without an OVR rating.

[Rumplebeanskin]

You would, but only if the engine used those stats as simply as A+B+C = specific number.

You could do the same thing with any game you wanted, any game that uses attributes. To me, overall tends to be a more at a glance number than anything else. If it's a reasonably important personnel decision, you'd be a bit silly to not consider a players attributes fully. Could go super hardcore and do away with overall ratings completely! It's not like you get to see them in real life

[Spharv2]

Yeah, even in Madden (ugh) it works fine. I played an online coach league in Madden '04 for years and had plenty of low rated guys who excelled for me because I targeted certain numbers, not OVR. I don't care if the game rates the guy a 43 OVR if he plays the way I want, I can plug him into the lineup and make a star out of him.

[goalieump413]

Well, since the overall rating is just an amalgamation of other ratings, including the system he's playing under, the overall rating should fluctuate from team to team.

But I have an idea: Instead of thinking in terms of overall rating being the barometer of a player's value, maybe a better way is to look at the overall as a measure of his capacity.

A lot has been made over the years about "current" and "potential" ratings, and how it's easy to gauge a player's value based on current and future ratings. In real life, players flame out for various reasons without ever reaching close to their potential. There are countless cases where 1st round draft picks only sign one contract, and are cut and out of football before their 2nd contract could come up.

I look at it this way: If a player has an overall rating of 84, let's say, then to me, his maximum value at his position is 84. That's not to say he's a "lock" at 84, it's just that he can only reach that rating value at the present time and not exceed it. He may train harder, get healthier, learn the system better, improve his attitude and ethic some too, which could then improve that value from 84 to something higher.

But he may also decline, meaning his max value is now in decline as well.

I'm not sure what the formulas have traditionally for games like Madden, but I believe it's a combination of the specific ratings based on position.

Also, it would be a good idea to track the ratings over time based on the incoming draft class and free agent marketplace. Free agent trackers often have ratings that help indicate positional strengths and weaknesses, but for an overall rating, again, not sure how the experts do it.

[Francis Cole]

Originally Posted by Rumplebeanskin

You would, but only if the engine used those stats as simply as A+B+C = specific number.

You could do the same thing with any game you wanted, any game that uses attributes. To me, overall tends to be a more at a glance number than anything else. If it's a reasonably important personnel decision, you'd be a bit silly to not consider a players attributes fully. Could go super hardcore and do away with overall ratings completely! It's not like you get to see them in real life

Currently they are not in and I don't know if I should add them. The post below yours said about how they were in Madden but not always right. If I copied that system you know people will file bug reports about it.

[Francis Cole]

Originally Posted by goalieump413

Well, since the overall rating is just an amalgamation of other ratings, including the system he's playing under, the overall rating should fluctuate from team to team.

But I have an idea: Instead of thinking in terms of overall rating being the barometer of a player's value, maybe a better way is to look at the overall as a measure of his capacity.

A lot has been made over the years about "current" and "potential" ratings, and how it's easy to gauge a player's value based on current and future ratings. In real life, players flame out for various reasons without ever reaching close to their potential. There are countless cases where 1st round draft picks only sign one contract, and are cut and out of football before their 2nd contract could come up.

I look at it this way: If a player has an overall rating of 84, let's say, then to me, his maximum value at his position is 84. That's not to say he's a "lock" at 84, it's just that he can only reach that rating value at the present time and not exceed it. He may train harder, get healthier, learn the system better, improve his attitude and ethic some too, which could then improve that value from 84 to something higher.

But he may also decline, meaning his max value is now in decline as well.

I'm not sure what the formulas have traditionally for games like Madden, but I believe it's a combination of the specific ratings based on position.

Also, it would be a good idea to track the ratings over time based on the incoming draft class and free agent marketplace. Free agent trackers often have ratings that help indicate positional strengths and weaknesses, but for an overall rating, again, not sure how the experts do it.

Potentials are in, but again it's not an overall potential it's a potential for every attribute. If we had the current/potential system (as detailed in other posts, completely different to OOTP based on the games/tapes your scouts watch etc), and then transform that into one overall grade, then again adjust that number into how your team plays I can see it getting very messy very quickly.

[TGH-Adfabre]

I liked the Madden Overall rating even though it was a simplistic sorting tool and not the most accurate assessment of talent. It sounds interesting to not have one though.

[Spharv2]

Yeah, I would be quite happy to not have one at all, or just go to a very simplified star or grade setup instead of a numeric thing. It's just a guide, and not a terribly useful one at that, aside from sorting purposes.

[Rumplebeanskin]

Originally Posted by Francis Cole

Potentials are in, but again it's not an overall potential it's a potential for every attribute. If we had the current/potential system (as detailed in other posts, completely different to OOTP based on the games/tapes your scouts watch etc), and then transform that into one overall grade, then again adjust that number into how your team plays I can see it getting very messy very quickly.

Agreed, and totally why it's a can of worms and not very simple to sort. Probably shouldn't have asked about it now

[goalieump413]

Originally Posted by Rumplebeanskin

Agreed, and totally why it's a can of worms and not very simple to sort. Probably shouldn't have asked about it now

I'll defer to Francis' response about potentials for attributes, as I think this is great news. But for the team a player plays for, I hope that eventually in v2 or beyond, that the effects of a team's play style and system modify these ratings, as coaches and GMs have to judge a player's value based on their system, not just a standardized system.

[dave1927p]

not sure how FOF does it but I enjoy their overall ratings generally.

[olivertheorem]

In general, I don't much mind there not being an overall rating. The one kicker, though, is that you lose the quick-and-dirty method of sorting players as a starting point. Even just a straight-up average of all ratings (while obviously not accurate and shouldn't be called an "overall" rating), would let you still have that.

[killershrew]

I like that there are no overall values. If you decide to go with overalls please put the option in to hide them. To take the discussion one step further there should only be exact values for measurable attributes, especially down the line when college football is available to coach/watch in the game.

A college scouting report should be very far from the insanely stupid madden scouting where you pay scouting points to see attributes. They should be more like this:
************************************************** *********
[insert combine stats here]
Joe Catchy, WR, 6'5" 210lbs

Pros: Joe is another long strider who just seems to glide past defenders. He is remarkably agile for a player of his size. He shows the ability to make people miss in the open field. Catchy is very tough to cover and has strong hands.

Cons: He takes plays off and has questionable work ethics. His ability to grasp complex schemes is average at best

[college stats]
************************************************** ********