|
calculated stats based on FINAL results
|
| TankNL |
Posted on 12-08-2014 04:13
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
ContadorVelasco wrote:
TankNL wrote:
ContadorVelasco wrote:
average results imo
I'm trying hard to appreciate you opinion, but can you explain yourself a little bit more? Did you do a test run? Are you replying on the results that OlegTinkov has posted? If so, what is average in you oppinion? Without that, you comment is rather useless, but that is just my honest opinion.
About Oleg results.
Chavanel only lost 3min to Nibali at stage with mountains and mountains.
im not have db, but im test here simulate stages and i see what results im have
What stage number are you referring to?
The cdb is downloadable from this link; https://dl.dropbo...lStats.cdb |
| |
|
|
| ContadorVelasco |
Posted on 12-08-2014 04:16
|
Under 23
Posts: 53
Joined: 27-07-2014
PCM$: 200.00
|
stage 10 |
| |
|
|
| zachbastian |
Posted on 12-08-2014 07:34
|
Free Agent
Posts: 131
Joined: 26-07-2010
PCM$: 200.00
|
must say that the test race by OlegTinkov looks promising! Good work TankNL |
| |
|
|
| TankNL |
Posted on 17-08-2014 22:15
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
I created a new matrix for the stats, file is attached here;
https://dl.dropbo...%20v4.xlsx
I also put the matrix in a database again;
https://dl.dropbo...lStats.cdb
Like last time i did a couple of adjustments on individual's;
Mountains;
Valverde 83 -> 81
Rodriguez 82 -> 81
Nibali 83 -> 82
Contador 81 -> 82
Froome 81 -> 82
Peraud 80 -> 81
ten Dam 78 -> 79 (I'm dutch  )
Hills;
Valverde 83 -> 81
Sagan 81 -> 79
Timetrail;
Tom Dumoulin 80 -> 82
Cancellara 80 -> 81
Cobbles
Cancellara 83 -> 82
Terpstra 83 -> 82
Sagan 81 -> 80
Sprints
Cavendish 82 -> 83
Sagan 83 -> 81
Fighter
Voigt 72 -> 80
Voeckler 76 -> 81
So only adjusted the stats on 14 riders, rest is all calculated based on actual results, from 1-1-2012 to 10-8-2014
I will do a update of the stats next week, just before the Vuelta.
Edited by TankNL on 17-08-2014 22:52
|
| |
|
|
| TankNL |
Posted on 18-08-2014 23:15
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
Adjusted the files including the .cdb with the races up until last weekend.
file in the links are updated;
Excel; https://dl.dropbo...%20v4.xlsx
Database; https://dl.dropbo...lStats.cdb
|
| |
|
|
| OlegTinkov |
Posted on 18-08-2014 23:47
|
Small Tour Specialist
Posts: 2666
Joined: 31-12-2007
PCM$: 450.00
|
Thanks!
|
| |
|
|
| TankNL |
Posted on 19-08-2014 00:14
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
I will create a new update on the 25th including;
| 18/08/2014 | | 24/08/2014 | | 2.HC | | Usa | | USA Pro Challenge - Tour of Colorado | | 19/08/2014 | | | | 1.1 | | Bel | | GP Stad Zottegem | | 19/08/2014 | | 22/08/2014 | | 2.1 | | Fra | | Tour du Limousin | | 22/08/2014 | | | | 1.1 | | Ned | | Arnhem - Veenendaal Classic | | 24/08/2014 | | | | 1.WT2 | | Ger | | Vattenfall Cyclassics - Hamburg | | 24/08/2014 | | | | 1.1 | | Fra | | Châteauroux Classic de l'Indre - Trophée Fenioux |
Edited by TankNL on 19-08-2014 00:15
|
| |
|
|
| TankNL |
Posted on 03-09-2014 19:28
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
For those who are interested, created a update with races up until this weekend and a special worksheet to see how the riders will evolve during the Vuelta; so stat changes due to the Vuelta.
https://dl.dropbo...%20v4.xlsx
If somebody is interested, I could put this in a database again.
Edited by TankNL on 03-09-2014 19:28
|
| |
|
|
| Ad Bot |
Posted on 22-11-2024 07:22
|
Bot Agent
Posts: Countless
Joined: 23.11.09
|
|
| IP: None |
|
|
| Macimate |
Posted on 03-09-2014 19:55
|
Free Agent
Posts: 125
Joined: 15-07-2008
PCM$: 200.00
|
Hey Tank, I remember you from the PPBD-forums. Nice to see veterans still having a crack at it.
Great concept by the way, I'm very interested in this. Would be cool if PCMdaily incorporated you into the next big release |
| |
|
|
| TankNL |
Posted on 03-09-2014 22:45
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
Macimate wrote:
Hey Tank, I remember you from the PPBD-forums. Nice to see veterans still having a crack at it.
Great concept by the way, I'm very interested in this. Would be cool if PCMdaily incorporated you into the next big release
Yup, have been in the "PCM world" from cycling manager 2 or 3
And joined the PPDB a long time ago... had some great discussion back then with Pimmen, who's the one and only stats-master 
But started with this project 3 years ago and every summer I add more code in order to make it better... still not entirely satisfied, but get the feeling that it is becoming better and better. Especially below the top 10/ top 20. I play a carreer with it and it play's really well. |
| |
|
|
| TankNL |
Posted on 04-09-2014 18:44
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
After heated discussions about the cobble stats of a number of riders, but especially Tim Wellens, I altered the point allocation for the GC in a WT2 stage race.
Now Tim Wellens has the following stats;
| CyclistName | FLA | MOU | HIL | TT | PRO | COB | SPR | ACC | DWN | FTR | STA | RES | REC | POP | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 69 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 | | previous; | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 56 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 |
Edited by TankNL on 04-09-2014 18:46
|
| |
|
|
| admirschleck |
Posted on 04-09-2014 18:48
|
Team Leader
Posts: 6690
Joined: 11-10-2010
PCM$: 200.00
|
That should be perfectly accurate. I agree that COB stats shouldn't be higher for the guys that just "survived" cobbles at certain races and won them afterwards, while of course, they should be for some guys that are specialists in some other aspects of cycling, but do good on them. Nibali is a great example.
Anyhow, I really like this system. Just keep it up!
|
| |
|
|
| matt17br |
Posted on 04-09-2014 18:52
|
Directeur Sportif
Posts: 10525
Joined: 28-09-2013
PCM$: 200.00
|
TankNL wrote:
After heated discussions about the cobble stats of a number of riders, but especially Tim Wellens, I altered the point allocation for the GC in a WT2 stage race.
Now Tim Wellens has the following stats;
| CyclistName | FLA | MOU | HIL | TT | PRO | COB | SPR | ACC | DWN | FTR | STA | RES | REC | POP | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 69 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 | | previous; | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 56 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 |
You are amazing. |
| |
|
|
| TankNL |
Posted on 04-09-2014 18:52
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
admirschleck wrote:
That should be perfectly accurate. I agree that COB stats shouldn't be higher for the guys that just "survived" cobbles at certain races and won them afterwards, while of course, they should be for some guys that are specialists in some other aspects of cycling, but do good on them. Nibali is a great example.
Anyhow, I really like this system. Just keep it up!
Thanks for the support. Due to the second place in the cobble stage of the TdF, Nibali does have a 73 in Cobbles, as does Fuglsang. Nothing manipulated there, just the points the gathered in a stage. Also Wiggins has a 73, due to his top10 place in Paris-Roubaix |
| |
|
|
| depeche92 |
Posted on 04-09-2014 19:20
|
Domestique
Posts: 489
Joined: 24-08-2009
PCM$: 200.00
|
TankNL wrote:
After heated discussions about the cobble stats of a number of riders, but especially Tim Wellens, I altered the point allocation for the GC in a WT2 stage race.
Now Tim Wellens has the following stats;
| CyclistName | FLA | MOU | HIL | TT | PRO | COB | SPR | ACC | DWN | FTR | STA | RES | REC | POP | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 69 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 | | previous; | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 56 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 |
His mountain stat is still wrong though. He should be 74 or 75
And also better Acc
Edited by depeche92 on 04-09-2014 19:22
|
| |
|
|
| TankNL |
Posted on 04-09-2014 19:28
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
have created 2 .cdb's, both based on PCMDailyEP 1.2
- 1 with the raw stats, no edit's; https://dl.dropbo...atsRaw.cdb
- 1 with a couple of changes for my own private liking; https://dl.dropbo...dStats.cdb |
| |
|
|
| TankNL |
Posted on 04-09-2014 19:46
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
depeche92 wrote:
TankNL wrote:
After heated discussions about the cobble stats of a number of riders, but especially Tim Wellens, I altered the point allocation for the GC in a WT2 stage race.
Now Tim Wellens has the following stats;
| CyclistName | FLA | MOU | HIL | TT | PRO | COB | SPR | ACC | DWN | FTR | STA | RES | REC | POP | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 69 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 | | previous; | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 56 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 |
His mountain stat is still wrong though. He should be 74 or 75
And also better Acc
kinda curious why you think that. Can't seem to find the results to support that. He did some breakaway's in the Giro and got nice results with them, but other than that? Personally I would give him high potential atleast 7, in order for him to grow to that level. |
| |
|
|
| depeche92 |
Posted on 04-09-2014 19:54
|
Domestique
Posts: 489
Joined: 24-08-2009
PCM$: 200.00
|
TankNL wrote:
depeche92 wrote:
TankNL wrote:
After heated discussions about the cobble stats of a number of riders, but especially Tim Wellens, I altered the point allocation for the GC in a WT2 stage race.
Now Tim Wellens has the following stats;
| CyclistName | FLA | MOU | HIL | TT | PRO | COB | SPR | ACC | DWN | FTR | STA | RES | REC | POP | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 69 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 | | previous; | | WELLENS Tim | 72 | 71 | 76 | 69 | 69 | 56 | 69 | 70 | 78 | 75 | 74 | 76 | 71 | 79 |
His mountain stat is still wrong though. He should be 74 or 75
And also better Acc
kinda curious why you think that. Can't seem to find the results to support that. He did some breakaway's in the Giro and got nice results with them, but other than that? Personally I would give him high potential atleast 7, in order for him to grow to that level.
He was 9th at Grappa MTT that is enough |
| |
|
|
| TankNL |
Posted on 04-09-2014 21:20
|
Domestique
Posts: 440
Joined: 19-03-2007
PCM$: 200.00
|
sorry, made error when editing....
here are the 1.2 files...
Raw stats; https://dl.dropbo...atsRaw.cdb
private edit; https://dl.dropbo...dStats.cdb |
| |
|
|
| Ulrich Ulriksen |
Posted on 06-09-2014 16:34
|
Directeur Sportif
Posts: 3262
Joined: 02-11-2010
PCM$: 300.00
|
Interesting work as always. Don't have '14 so I can't test for you. But will get it at some point.
Two ideas:
1. Not sure if/how you resolved the issue of young improving or old declining riders getting earlier periods over-weighted. But the thought I had was that you could look at their base results and then use a different weighting of the years based on the trend. If the rider had improved from 2012 to 2013 to 2014 by more than X%each year you flag them as improving and overweight 2014, same if they got consistently worse. But if they bounced around you use more of a straight average of the years. This would also avoid penalizing a rider who was injured in the current year too much or rewarding a rider who had one great year too much.
2. Would be cool to combine this with my fictionalized database. Instead of me generating randomized PCM stats I would create 3 years of race results based on general rider types. Then run them through your algorithm to generate game ratings. This would have two advantages over my current process (1) would benefit from the great stat stability in your system and (2) it would create history so that when you start a game you have 3 years of history built in. A lot of challenges to this in creating good results history but something for me to think about. Does your system rely only on stage results or does it also factor in the overall Tour results for stage races? |
| |
|
