hi folks! now i know the PCM community is tight and alot of players have put some serious work into making our favorite game better and more realistic, BUT there is one thing that has been bugging me. A long time ago a list of stat effects was circulating on these forums and many people took that list as being the truth. I am going to try and debunk the myths that have stemmed from that list.
Flat: indicates how a rider will go on the flats and moderate descents...easy
Mountains: indicates how a rider will go on climbs that are steady and long and will influence the performance on TTs with climbs. Any incline over roughly 10km is considered a Mountain.
Hills: indicates how a rider will go on shorter climbs or undulating hills. the difference between a mountain and a hill is like i said about 10km. if a longer climb is broken up with easier grades it can have the effect of a serious of hills as apposed to one mountain climb.
TT: indicates a riders performance in TTs over 20kms that are flat. If a TT involves a large amount of climbing than the Mountains stat will come into play.
Prologue: indicates a riders ability in TTs under the 20km mark
Cobbles: cobbles drain energy from riders and so the more of this stat a rider possesses the less of an effect the cobbles will have on his race. depending on if the cobbles are flat or hilly will further influence the riders effort.
Sprint: indicates the top speed the rider can achieve while sprinting or attacking. no more, no less.
Acceleration: indicates the rate at which a rider achieves his top speed during a sprint or attack. no more, no less. (this stat is often said to influence the amount of "red energy" your rider has. this is 100% FALSE.)
Downhill: indicates a riders top speed and ability to stay upright on longer, steeper descents. (this oft-overlooked stat allows riders with high DH ability to out pace rivals by long continued attacking down steep slopes. A riders energy is virtually untouched by attacking in the descents and can create up to 20km/h difference in rider speeds. The likely-hood of a rider crashing on a fast descent is also greatly reduced is their DH stat is high.
Fighter: while working in a breakaway a rider will experience a similar energy sapping effect as to riding on cobbles. the very act of being away from the peleton drains on the riders. having a high Fighter stat will work to counteract this effect just as a high Cobbles stat will counteract the rough stuff. this is easily proven with Jens Voigt, whos 85 Fighter skill allows him to preform virtually the same whether alone, with a small group, or in the bunch. (NOTE: this stat DOES NOT effect how often a rider will attack if controlled by AI.)
Stamina: this is the stat that will determine the rate at which your "red energy" is drained as compared to other riders. this stat also will influence your "yellow energy" but many other stats also influence the "yellow" energy (such as being in a breakaway)
Resistance: this stat indicates how well the rider will deal with longer race distances. it is a sliding scale with the highest Resistance riders being able to preform consistently over 200km+. This is the stat that will have the most influence on your "green energy"
Recovery: indicates how well a rider will recover from day to day efforts during a stage race. (despite repeated testing i am still unable to say with certainty if this stat influences how well a rider will recover from efforts DURING a race)
thanks for reading and please feel free to ask questions or insult me!
top 10 of all time in no particular order, not including the cannibal cause hes automatically #1:
-Gemininani -Nencini -Anquetil -Coppi
-Kelly -Hinault -Zabel -Bettini -Rebellin -Magni
..and honorable mention to Tom Simpson whos career was cut short.
Top 5 noobs most likely to tear it up this season:
-Phinney -Moser -TJ -Guardini -DURBO!
I definitely won't insult you, but sobrano did excessive testing as well and came to completely different results in acceleration, fighter, stamina and resistance (with the latter two obviously switched in your text).
see thats exactly why i did my testing! so many people had what i figured was the "wrong idea" so i did what has amounted to months of trial and error. isolating external factors was the hardest part and just requires alot of patience as using anything but the real race mode is useless.
basically i looked at what sobrano did and tried to replicate his results. what i got was radically different so i took it a step further and just did the tests my self. riders like Jens, Frank Schleck, and Cancellara were vital as they showed extremes in stats.
the most controversial is the stamina stat im sure. its an easy test. go play with Freire and Frank Schleck and youll see clear as day at ACC actually means ACC haha.
top 10 of all time in no particular order, not including the cannibal cause hes automatically #1:
-Gemininani -Nencini -Anquetil -Coppi
-Kelly -Hinault -Zabel -Bettini -Rebellin -Magni
..and honorable mention to Tom Simpson whos career was cut short.
Top 5 noobs most likely to tear it up this season:
-Phinney -Moser -TJ -Guardini -DURBO!
Well, even though I have limited knowledge of stat testing, I'm sure you did one mistake that might explain everything:
When testing the effect of a single stat, it's very important that the two riders you compare are the exact same apart from this stat, and that previous efforts are exactly the same.
From what you've written i've got one question now
did you ever change any riders stats when testing? ie. set all stats of several riders to 60 except the stat you were testing
TheManxMissile wrote:
From what you've written i've got one question now
did you ever change any riders stats when testing? ie. set all stats of several riders to 60 except the stat you were testing
oh yes of course! thats the kind of testing i tried to replicate to get sobranos results. you would think that isolating stats like that would be the best option but i found making all the riders decent at everything (at around 70) and then testing stats at 85 was alot easier.
the logical fallacy there is that by reducing the other stats you create a pure comparison. sadly the game doesnt work like that and creating riders at 55 but with one or two stats at 85 produces strange results. that led me to believe that most (if not all) the stats are interconnected in some way and certainly much more connected than we had originally thought.
Edited by CosmicOsmo on 30-01-2013 00:20
top 10 of all time in no particular order, not including the cannibal cause hes automatically #1:
-Gemininani -Nencini -Anquetil -Coppi
-Kelly -Hinault -Zabel -Bettini -Rebellin -Magni
..and honorable mention to Tom Simpson whos career was cut short.
Top 5 noobs most likely to tear it up this season:
-Phinney -Moser -TJ -Guardini -DURBO!
cio93 wrote:
Well, even though I have limited knowledge of stat testing, I'm sure you did one mistake that might explain everything:
When testing the effect of a single stat, it's very important that the two riders you compare are the exact same apart from this stat, and that previous efforts are exactly the same.
i did not make that mistake haha. in fact it was the first thought i had before i began testing
top 10 of all time in no particular order, not including the cannibal cause hes automatically #1:
-Gemininani -Nencini -Anquetil -Coppi
-Kelly -Hinault -Zabel -Bettini -Rebellin -Magni
..and honorable mention to Tom Simpson whos career was cut short.
Top 5 noobs most likely to tear it up this season:
-Phinney -Moser -TJ -Guardini -DURBO!
Well, some phrases made me believe you didn't, basically any containing a rider name, as that implies you left them with their characteristic skill sets.
Still, I doubt stamina and resistance have effects one would attribute to the respectively other one purely based on their names, and then we have testimony against testimony regarding fighter and acceleration.
And while we know that stats interconnect in some ways (FL+TT+RES for TT performances f.ex.), the tested stat wouldn't change its effect.
55/55/.../85 and 55/55/.../55 should differ by the exact amount of influence the last stat has on the tested efforts. What are the strange results you observed?
Edited by cio93 on 30-01-2013 00:30
the named riders and their performances are what tipped me off to thinking sobrano was wrong. and i did testing with real rider profiles and my own made up ones. like i said, thinking that testing at 55/85 will yield accurate results is a logical fallacy simply due to how the games engine works. if you dont believe me simply do the tests yourself. thats what i was hoping for. that others would start looking at this for themselves instead of just parroting what one person came up with SEVERAL versions ago.
the fighter stat is the EASIEST to test. so go try it. youll see that the riders with the highest fighter stats dont attack any more frequently than riders with no ability in that area.
the resistance test is also easy. get a stage with 200km+ line up a peleton of riders at 70 base line. then adjust the resistance levels from 55-85. next step is creating a scientific environment. so that means keeping the peleton at a set speed (easiest on flat stages) and removing effects like weather, temp preferences, and fitness. then watch as the green bars drain. riders with no resistance will lose green faster than those with. remember you need lots of control on the stage and lots of subjects incase u get riders flatting, crashing, or what have you.
what i said about schleck and freire is valid. look at the stats and then play with them on your own. its an easy way to see how those stats work.
also from a semantic point of view: what you consider to be a logical definition of what "stamina"should be is irrelevant. A, the game is not made in english. B, we all define words differently.
top 10 of all time in no particular order, not including the cannibal cause hes automatically #1:
-Gemininani -Nencini -Anquetil -Coppi
-Kelly -Hinault -Zabel -Bettini -Rebellin -Magni
..and honorable mention to Tom Simpson whos career was cut short.
Top 5 noobs most likely to tear it up this season:
-Phinney -Moser -TJ -Guardini -DURBO!
Stamina is deffinatly what you explained as resistance. The over 200 km thing. No need to do any stat testing for that. Have played Milan - San Remo over 100 times now and It is almost impossible to get a good result with a rider with low stamina.
Spilak23 wrote:
Stamina is deffinatly what you explained as resistance. The over 200 km thing. No need to do any stat testing for that. Have played Milan - San Remo over 100 times now and It is almost impossible to get a good result with a rider with low stamina.
while your observation is appreciated you are falling victim to what others have said is an easy mistake, your comparing riders who in all likelyhood have high resistance and stamina. IN FACT i seem to remember a tool tip on the PCM loading screen even saying that resistance is the distance stat. id stake my race bike on that. but please do the test yourself and see. also look at the riders who are known as "classic specialists" or at past Milan san Remo winners. all have high resistance ( and most likely high stam too)
top 10 of all time in no particular order, not including the cannibal cause hes automatically #1:
-Gemininani -Nencini -Anquetil -Coppi
-Kelly -Hinault -Zabel -Bettini -Rebellin -Magni
..and honorable mention to Tom Simpson whos career was cut short.
Top 5 noobs most likely to tear it up this season:
-Phinney -Moser -TJ -Guardini -DURBO!
I btw. never read sobrano`s thread but can definitely speak from my own experience. And to be sure I just tested it.
I created a DB with all riders having 60 in every single stat.
2 teams were highered to 85 STA.
2 teams were lowered to 50 STA.
2 teams were highered to 85 RES.
2 teams were lowered to 50 RES.
Then I played Milan San Remo. (Cyanide set 14 riders with 85 STA to the favourites in the loading screen btw.)
Result was that the 85 STA teams did the pace for all the 299km.
When the climbing part on the final 30km begun 13! riders couldn`t follow the pace in the peloton. All (!) of them were 50 STA riders.
The final sprint was won by a 85 STA rider. The top 10 had 8 85 STA riders. the 10th was a 60 everywhere rider. And a bit surprising the 3rd was a rider with 60 STA and everywhere else BUT with 50 RES.
So I highly doubt that your calculation is correct when even a 50 RES rider can sprint to 3rd next to all the 85 STA guys while 50 STA guys finish minutes later and no 85 RES rider had any impact on the race.
CosmicOsmo wrote:
Any incline over roughly 10km is considered a Mountain.
[...]
the difference between a mountain and a hill is like i said about 10km.
[...]
TT: indicates a riders performance in TTs over 20kms that are flat. If a TT involves a large amount of climbing than the Mountains stat will come into play.
Prologue: indicates a riders ability in TTs under the 20km mark
You are talking about us having some "concepts" we consider the truth - well, if you criticize that (and I agree you should if you disagree with it!), don't put out numbers like they were the absolute truth either
Length, slope, constancy, or whatever of these doesn't influence whether the used stat is Mountain or Hill. What defines this is a value called gene_i_mountain, which up to PCM11 was set in the DB and in PCM12 is read by the game directly from the stage. Of course, the variables I mentioned above influence on the gene_i_mountain value of each stage. But once the value is set, it's the same for all climbs. So if you have a stage with 3 .HC climbs, and a couple of 1km hills somewhere, very likely the stage will have a value of 1 in gene_i_mountain which means that the MO value will be the one used in those hills as well.
On Prologue and Time trial, most times it's a mix of both. That was confirmed by someone working with the game iirc. According to the testing Daniano did in PCM10, the 'weight' of each stat according to lenght of the TT is approx.:
5km = 100% Prologue
10km = 80% Prologue & 20% TT
15km = 60% Prologue & 40% TT
20km = 40% Prologue & 60% TT
25km = 20% Prologue & 80% TT
30+km = 100% TT
I've once tested acceleration to see what happened, it was exactly as what you say it isn't, the red bar took longer to be decreased in the rider with higher ACC. But I'll later re-test it, with the stats you're suggesting to try.
I can confirm that all riders accelerate (in the traditional sense of the word) the same.
I created a database with (only) 4 riders, all with 65 stats for everything. I then changed the acceleration of 1 of those riders to 55, 1 of those riders to 75 and 1 of those riders to 85. I created a custom stage that was only 25 k's long, and pancake flat. I kept the riders together, and had them all sprint after the 3 k mark. No rider was left behind until the red bar was depleted. The riders mostly got dropped in the same order, no matter the position of where they started; 55 first, 65 second, 75 third and 85 always the last one left of the 4. The test was repeated 11 times. There was an occasional variance in results, but riders were only ever +/- one position.
So I highly doubt that your calculation is correct when even a 50 RES rider can sprint to 3rd next to all the 85 STA guys while 50 STA guys finish minutes later and no 85 RES rider had any impact on the race.
like i said you CANNOT use min/max stats and get reliable results. use a baseline 70 and repeat the experiment. like i said the game engine wont allow for accurate results when you use min/max.
as for the mountains, i am aware that the game designates mountains in the code but riders with high hill stats will consistently perform on climbs labeled "mountain" if the incline varies from high to low. set a rider at base 70, give him 55 in the mountains and 85 in the hills. now pick a good climb...some of the ones in romandie or swiss would work....that has varying incline. your gilbert-esque rider will be great on the climbs despite no mountain skills.
i want to say thanks to everyone passionate enough to look into this sort of thing and willing to help gather data. look if im 100% wrong and you are all getting sobranos results then i gotta keep trying but my goal here was to shake those convictions up a bit. my proof was in the pudding, after my new understanding of stats i have pretty much mastered the game on extreme and theres not much left that challenges me so working these stats out has been fun.
top 10 of all time in no particular order, not including the cannibal cause hes automatically #1:
-Gemininani -Nencini -Anquetil -Coppi
-Kelly -Hinault -Zabel -Bettini -Rebellin -Magni
..and honorable mention to Tom Simpson whos career was cut short.
Top 5 noobs most likely to tear it up this season:
-Phinney -Moser -TJ -Guardini -DURBO!
Ok, here are my resistance, stamina, sprint, acceleration and figther tests:
First of all, disabled all variables ofc. No daily form, no random form, no country bonus. For the first tests, I got Cavendish, T.Martin and Boonen with 70 in all stats. To avoid working with extremes, I used three riders: two in the extremes, one in the middle (85,70,50).
Test 1
First test was with stamina. With all other stats 70, I gave Cavendish 85 STA, Martin 70 STA and Boonen 50 STA. The test was to let them drop off the peloton several minutes, put a 10~15 sec gap between them (just enough so that each are in their own group, although I'm not sure whether this would actually affect it) and let them ride dot 50, later 60, all the way in Milan San Remo.
In the beginning, the gap between them remained the same. When they finished descending the Passo del Turchino, I upped their effort to 60. Some kilometers later, Boonen started to be left behind. https://img208.imageshack.us/img208/37...627463.jpg
Notice that on the screens, their heart rate and green bar gone is always the same - what changes is the efficiency of the energy spent.
Test 2
Second test was now with resistance. All other stats at 70, Cav with 85 RES, Martin 70 RES, Boonen 50 RES. Set up the same 10~15 seconds gap and let them ride, same as above.
Test 3
Ok, resistance definitely does not affect the distance. But is it true then that it affects yellow bar comsuption?
With the same setup as test 2, the beginning procedure was the same: let all riders drop from the peloton. No need to create gaps between them this time, though. Now, let's put them in a yellow-bar comsuption effort. At first I set dot 70, but soon upped to 80.
The gap between them never got really big in the next kms (about 20 seconds), that is, until Tony Martin fell and got up 2 minutes behind Cav (what are the odds )
Test 5
Now the acceleration test. I mean, this has been tested dozens of times by different people, but one more won't hurt. Same setup as always, now Cav 85 ACC, Martin 70 and Boonen 50.
Test 6
Finally, the fighter test! This was the one that required most work on the setup. What I basically did was: a TdF peloton where all riders have 70 on every stat, except for a team leader for all teams which have the same stats; from the remaining 8 riders, 2 have 85 FTR and the other 6 have 50 FTR (yes I know this follows the "extremes" case but I don't think it would affect this one particularly).
They're all set to ride a flat TdF, and of course what we're looking to see here is which riders will go in breakaways.
At first, most riders who attacked to form breakaways had 50 FTR; from about 20 riders who attacked, 2 or 3 only had 85 FTR I think. So I was (actually I am) inclined to agree with you here. However, as I wished for more attacks, I set my entire team to relay and brought back the breakaway. Now at about halfway of the stage, 8 more riders attacked, all of them with 85 FTR! After a while, I brought them back; 6 more riders attacked, all with 85 FTR, until finally a 50 FTR rider followed, together with other 2 max. FTR riders.
So that was 14 consecutive 85 FTR attacks; considering that leaders wouldn't attack, if I have my math right the probability of 14 consecutive attacks by riders with 85 FTR are of 0,0000037%. So that leads me to believe that the FTR stat does influence the attack by AI riders, but it depends on the stage situation; if I had to guess, halfway through the stage the likeliness of attacks by high FTR riders are definitely higher, but in the beginning of the stage it has little effect. The rest could also work the way you've described (less energy comsumption), but well, I don't think we've ever heard such explanation (keep in mind that testing isn't the only source we got for info about the stats; some of these explanations came from people who actually work with the game).
And I guess that about wraps about it up. Another funny thing I tested earlier but wasn't thinking it was worth a comment: while taking a quick look at one of sobrano's stats explanation thread, I read that in TTs, the TT/PRL mix replaced the flat stat. Well, I tested three riders (guess who ), one with 50 FL and the other with 85 and the same on all other stats, guess what? They rode home in the same time! So don't come again saying that even if Kelderman [undeservedly ] had 76 TT, he'd be behind Castroviejo because of his secondary stats = 7 lower points of flat, Jesleyh. Flat has no weight on time trials!
![imgur.com/KlJUqDx.jpg[/img[img]]https://imgur.com/yUygrQ.jpg](https://imgur.com/KlJUqDx.jpg[/img[img]]https://imgur.com/yUygrQ.jpg)
)
