|
Moving to PCM22
|
|
| Ad Bot |
Posted on 21-11-2024 19:43
|
Bot Agent
Posts: Countless
Joined: 23.11.09
|
|
| IP: None |
|
|
| Ollfardh |
Posted on 10-11-2023 12:56
|
World Champion
Posts: 14562
Joined: 08-08-2011
PCM$: 9100.00
|
I didn't want to rant in the race threads, as I have an immense respect for the reporters and they don't deserve the resulting negativity, but now that's it's here..
WTF is this season. I was really hoping PCM22 would mean a step forward but it's 3 steps back. Classics are ok, but stage races are a disaster. Sprinter races are the worst, both in PT and PCT. Juarez was just criminal what happened to Groenewegen.
For my won team Bonifazio's performance unfortunately didn't improve in PCM22, but I can live with that. It's the stuff I can't explain like Oomen sitting in the back and losing 2 minutes after a great prologue in Ukraine. Or Novak losing 3 minutes as one of the best climbers in Portugal.
With Cio's analysis of PCM23, I don't see a bright future for the MG at this point.
Changed my sig, this was getting absurd.
|
| |
|
|
| ivaneurope |
Posted on 10-11-2023 14:22
|
Classics Specialist
Posts: 2933
Joined: 09-05-2011
PCM$: 300.00
|
With the risk of repeating myself - I think going back to an older version will send the wrong messeage. It could alienate existing managers from the MG and also discourage potential new managers and reporters. But maybe I'm not speaking from an experience as I'm not a reporter (due to lack of PCM22 and given that Abhi had issues, I don't think my PC will sustain it even if I had a copy) and never had been one. I'm speaking more from a participant's point of view and I believe that for now sticking with PCM22 for a while would be the wiser choice. Given Cio's analysis of PCM23's AI behaviour, a move to this version doesn't seem teanable, but if the behaviour of the AI in the future PCM24 is better, perhaps we should discuss using it for the 2025 season onwards.
As for the seasons with the PCM22 engine, I assume that teams will be building more TT-heavy rosters considering that the TT stats are (at least from what I've read here) massively OP'ed
|
| |
|
|
| SotD |
Posted on 10-11-2023 14:44
|
World Champion
Posts: 12187
Joined: 29-11-2006
PCM$: 2980.00
|
The worst possible thing we could do for the game is to shut our eyes for the sake of moving forwards. A step back definately isn’t catastrophic if that is what give the best possible results for the MG.
|
| |
|
|
| whitejersey |
Posted on 10-11-2023 14:45
|
Classics Specialist
Posts: 2904
Joined: 07-08-2011
PCM$: 300.00
|
Ollfardh wrote:
I didn't want to rant in the race threads, as I have an immense respect for the reporters and they don't deserve the resulting negativity, but now that's it's here..
WTF is this season. I was really hoping PCM22 would mean a step forward but it's 3 steps back. Classics are ok, but stage races are a disaster. Sprinter races are the worst, both in PT and PCT. Juarez was just criminal what happened to Groenewegen.
For my won team Bonifazio's performance unfortunately didn't improve in PCM22, but I can live with that. It's the stuff I can't explain like Oomen sitting in the back and losing 2 minutes after a great prologue in Ukraine. Or Novak losing 3 minutes as one of the best climbers in Portugal.
With Cio's analysis of PCM23, I don't see a bright future for the MG at this point.
Just saying this isn't exactly new to PCM22, this also happened to LLS and I last year in a stage race where Jakobsen and Kennaugh decided they didnt want to sprint. Sprints have always been a massive crapshoot in PCM, also if there is any sprinter that has been royally screwed this year it's Coquard in PT.
|
| |
|
|
| ivaneurope |
Posted on 10-11-2023 15:18
|
Classics Specialist
Posts: 2933
Joined: 09-05-2011
PCM$: 300.00
|
SotD wrote:
The worst possible thing we could do for the game is to shut our eyes for the sake of moving forwards. A step back definately isn’t catastrophic if that is what give the best possible results for the MG.
On the contrary there won't be that many reporters. And this totally excludes new reporters from the equation as they will need to own a copy of the older game. I'm honestly against moving down to an older version - it doesn't seem right for me. And while I'm saying this from a participant POV and not a reporter I'm not sure we can get new managers (and new reporters) or keep existing managers interested in MG should we go 2 steps back to get 1 forward.
It's a debate that had been raging on for...as long as I've been around in the ManGame. Quite frankly, I don't see an end to it.
|
| |
|
|
| SotD |
Posted on 10-11-2023 15:40
|
World Champion
Posts: 12187
Joined: 29-11-2006
PCM$: 2980.00
|
I think moving forward is more expensive in terms of buying games, that moving backwards, so I really don’t see that as a factor. Especially not as me and a couple of others offer to pay for the games every season.
I really don’t think anyone participates in the Man-Game because of the PCM engine used. It would be a very Strange incitament for at game where your only participation is based on everything else than the racing itself.
|
| |
|
|
| alexkr00 |
Posted on 10-11-2023 15:43
|
World Champion
Posts: 13915
Joined: 05-08-2008
PCM$: 300.00
|
I've always been an advocate for always moving to newer games, but I'm not a big fan of what I've seen so far.
Yes, sprints have always been kind of random, but it feels like it's much worse this year with top sprinters constantly failing to deliver. And it's not just one. Or one sprinter being ignored in a race. That's not new. It feels like all of the top sprinters have had more misses that hits so far.
Anyway, I'd wait until the end of the season and have a look at the points tally of riders before deciding if the new game is suited or not for the MG.
|
| |
|
|
| Eden95 |
Posted on 11-11-2023 11:45
|
Grand Tour Specialist
Posts: 4505
Joined: 05-10-2010
PCM$: 900.00
|
Agree with Alex. Sprints have always been problematic, but usually the burden is shared throughout the season and it’s pretty inconsistent. However this season it’s been consistent through several races now that the top sprinters are basically sabotaged by the game engine. The top 2-3 favourites will always have leadouts (like in Juarez it was Fastned and Assa every stage) that will almost always fail. It’s genuinely detrimental to a team to have a great sprinter on their squad right now, and unless this miraculously changes those teams are going to massively under perform and be at risk of relegation/not be in promotion contention as a result which is totally unfair. Honestly, having an average (or at least not top of the line) 80-81 sprinter and a decent 78-79 sprinter with good FL/energy stats is probably the play right now (in PCT) if you want results in sprint races as they’re much less likely to be considered a top 3 favourite by then game instead of a big name and wage rider which isn’t right.
Indosat - ANZ HQ
"This Schleck sandwich is going to cause serious indigestion for Evans" - Phil Liggett |
| |
|
|
| DubbelDekker |
Posted on 11-11-2023 14:09
|
Small Tour Specialist
Posts: 2633
Joined: 20-04-2008
PCM$: 200.00
|
Maybe we can continue using the newest engine for the types of races that it has acceptable performance in and run the races that get destroyed by it in an older engine?
As long as everyone is aware pre-transfers I don't see any problems with fairness.
But maybe it adds too much complexity for organizers and reporters?
|
| |
|
|
| ivaneurope |
Posted on 11-11-2023 16:57
|
Classics Specialist
Posts: 2933
Joined: 09-05-2011
PCM$: 300.00
|
DubbelDekker wrote:
Maybe we can continue using the newest engine for the types of races that it has acceptable performance in and run the races that get destroyed by it in an older engine?
As long as everyone is aware pre-transfers I don't see any problems with fairness.
But maybe it adds too much complexity for organizers and reporters?
It will become too messy especially for reporters, who may have to own copies of both engines (e.g. PCM20 and 22). I'm not sure current reporters would agree and I'm pretty certain that potential new reporters will not be interested in this arrangement since they may not own PCM20 (don't think if it's even available as game studios tend to end support for older titles, but I could be wrong).
Then again, I'm speaking only from a participant's viewpoint and my oppinion has less weight to a reporter who has reported on races on both versions. I'm sounding like a broken record with this, but I'm against going backwards (even partially to mitigate anomalies). However, this stance of mine will not affect my commitment to the ManGame and will accept any decision made by the MG's organizers.
SotD wrote:
I think moving forward is more expensive in terms of buying games, that moving backwards, so I really don’t see that as a factor. Especially not as me and a couple of others offer to pay for the games every season.
I really don’t think anyone participates in the Man-Game because of the PCM engine used. It would be a very Strange incitament for at game where your only participation is based on everything else than the racing itself.
I've considered joining in as a reporter, but I've decided against it. Don't have PCM20 or 22 and it doesn't feel right for me to have something that I don't technically own. But that's a whole other topic that I don't want to discuss here.
|
| |
|
|
| SotD |
Posted on 11-11-2023 17:21
|
World Champion
Posts: 12187
Joined: 29-11-2006
PCM$: 2980.00
|
Ivan, why do you insist on taking on the point of view from the reporters, when you obviously aren’t interested in that role? It’s pointless to keep on making assumptions regarding something you don’t have neither experience with nor interest in.
It’s definately NOT a problem to have access to multiple versions at the same time. It might be an issue to swap between versions, but let us leave that response for people actually active reporting.
Regarding moving backwards we might face an issue of having only a certain amount of stages to pick from - but in all honesty - is that really a problem? Is anyone playing the Man-Game because of boostes graphics or a huge variation in stages in races? I would ve very surprised if that is a point that keep people in the game.
|
| |
|
|
| Ulrich Ulriksen |
Posted on 11-11-2023 20:15
|
Directeur Sportif
Posts: 3259
Joined: 02-11-2010
PCM$: 300.00
|
Not sure I know the answer on versions but in terms of solutions:
I don't see a problem reporting on two versions, if a reporter only had one version they just do the kind of race that is done on that version. It actually might open up more reporters since you just have to have one version or the other to do some races.
The bigger work is for the DB team - having to create duplicate versions of the DB.
Because stages aren't backward compatible it might create some complexity but would agree that isn't a big limitations.
Man Game: McCormick Pro Cycling |
| |
|
|
| jph27 |
Posted on 11-11-2023 20:39
|
Team Leader
Posts: 7339
Joined: 20-03-2010
PCM$: 900.00
|
Given we've previously used PCM 18 and PCM 20, I think reporters not having the game may not be a huge issue? I think most current reporters have copies, and there are others such as myself who would (time permitting) be happy to help out with reporting but don't have PCM 22. From a quick check of Steam there isn't an issue with availability either, unless we go much further back - PCM 17 is the last I can find there.
Part of the challenge in working out the best way forward is that it's not clear how much of this is an AI issue, and how much is an issue with the Man-Game DB. Not just thinking about the number of riders with 80+ main stats, but also considering riders were added with stats designed for the game version in use at the time, and now the way those stats influence the AI have changed. There may not be an easy fix for that, but it may explain some of the more bizarre AI we see.
There is potentially some value in running tests to better understand how the different stats influence the AI, to see what mitigations might work, but that's likely to be complicated and time intensive, unless anyone has ideas about a simplified way to do it? TTs are easy enough to model the influence of different stats in, but other stages are trickier - and TTs seem to work now anyway! |
| |
|
|
| Fabianski |
Posted on 11-11-2023 21:48
|
Grand Tour Specialist
Posts: 4667
Joined: 29-09-2018
PCM$: 185.00
|
@Ivan, I'll happily give you some insights from a reporter point of view. Although this is just me, every reporter obviously has his own thoughts - I don't think there's THE point of view from reporters.
I started reporting in my first MG season, 2019. Which was on PCM18, just like our 2020 season. It's a game version I've been playing hours and hours with, even outside from MG. I did like most aspects of it, while there obviously were some downsides as well.
In 2021, when we moved to PCM20, I didn't do any reports. The first results of the season just killed my motivation, and it didn't get that much better over the year. Especially sprinter stage races were (still) broken, with the known GC leader and 4th sprinter bugs. In 2022, I decided to do some reports again, just because it was needed for MG to keep going. And I think the results weren't that awful generally speaking, I just hated reporting sprints, knowing that some sprinter(s) will again randomly miss out. I feel like this was less the case in PCM18, but I think it happened there as well. And btw, I didn't play the game a single minute outside of MG races.
Now, in 2023, I got my PCM22 copy - but looking at the results so far, especially in stage races, it's pretty clear for me that I won't do any reports with that version. I just hate to disappoint managers (which is my own problem, sure), and even more so managers of top riders who simply don't perform.
Long story short: at least in my case, it's not the newer game version that gives me more motivation for reporting - it's actually 100% the other way round right now.
But yeah, I'm also wondering what kinds of approaches we could take. Cio's experience suggests that sprints are even worse in PCM23, with even favourites who are lightyears above the rest not participating. At least for the sprints issue, PCM23 won't be the solution, sadly.
And going back is no option either for half of the managers (and yeah, I'm fully in line with alex, always attempting to take the next version, but I'm no longer there).
Having two versions in parallel probably means like twice as much work for the DB team to get everything ready. It's not really a viable option, either - but for now, it could be the only one that gives us at least partially better results in certain types of races.
But we'll need to observe the AI behaviour throughout the 2023 season first, and maybe try to come up with solutions after TDF or so. For me, staying with PCM22 is not a viable option right now - but it's 100% sure that there won't be a solution everyone will be happy with. Up to MGUCI to make that (difficult) choice.
|
| |
|
|
| whitejersey |
Posted on 11-11-2023 22:23
|
Classics Specialist
Posts: 2904
Joined: 07-08-2011
PCM$: 300.00
|
I personally think that we try to stay with PCM22 and then we will have to find out how to make some races run better. As we have seen plenty of times multiple hills races have been hampered by the fact that the TTs have been too dominant. I still think that a lot of the issues around hills stages can be sorted in calendar creation, like just removing some of the TTs from races that really shouldn't have them.
Sprints won't become magically better if we switch back editions of PCM, as the 4th sprinter issue will exist then. No matter how we spin it I feel like there hasn't been any editions of PCM that actually does AI trains well enough to warrant it. As I pointed out before I got royally fucked last year by sprint AI on a few occasions with Kennaugh deciding that he did not want to participate in sprints when he should have.
Just looked at the reports for Juarez, and this looks exactly like what happened to LLS and I last year in Guadeloupe(IIRC) and we didn't have people asking for an edition change but we got a few sorrys, Just like when PCM was unable to handle my amount of strong puncheurs in my first PCT season, it was a mistake on my part for building my roster in that way etc. I AM NOT SAYING THE MANAGERS IN QUESTION MADE MISTAKES. But looking at the energy stats of the leadout in Vingerling and Matthews almost no matter how you spin it Groenewegen would almost always be delivered too early because they run out of energy fightiing for positions, I am not saying that trains are perfect but trying to figure out what makes a great PCM leadout could be a way to fix it going forward.
As TMM have been saying for years the DB is the biggest problem by a longshot but I can understand people being attached and not wanting to nuke it, which might be the best option of a hoist of bad options. A lot of the paths for riders feel so lackluster since they were designed for previous PCMs so we're bound to have some holes.
Also we're 2,5 months into the season, I think that people are currently overreacting because something is not what it used to be, the early signs are somewhat dooming but we got to realise that there is still almost a full season to be run. The longer we commit to a specific version of the game the more we will be able to iron out these kinks.
I will say that I have been travelling for close to three weeks and have been ill for a week so I am not fully ingrained into what's going on results wise across every single division, but I don't think that going backwards is going to be the answer.
|
| |
|
|
| TheManxMissile |
Posted on 11-11-2023 22:28
|
Tour de France Champion
Posts: 18187
Joined: 12-05-2012
PCM$: 0.00
|
whitejersey wrote:
As TMM have been saying for years the DB is the biggest problem by a longshot but I can understand people being attached and not wanting to nuke it, which might be the best option of a hoist of bad options. A lot of the paths for riders feel so lackluster since they were designed for previous PCMs so we're bound to have some holes.
*quielty whispers "soft reboot" into the wind and runs away*
|
| |
|
|
| Laurens147 |
Posted on 13-11-2023 07:50
|
Protected Rider
Posts: 1122
Joined: 26-04-2020
PCM$: 800.00
|
In relation to the sprint issue: what is strange to me is that Milano-San Remo was almost a flawless sprint result. Of the 11 favorites I listed in the preview, 9 favorites finished in the top-10. So it should be possible one way or another?
MG - Lotto - Caloi
[MG] New Manager of the Year - PCM.daily Awards 2022
|
| |
|
|
| Fabianski |
Posted on 13-11-2023 07:59
|
Grand Tour Specialist
Posts: 4667
Joined: 29-09-2018
PCM$: 185.00
|
Laurens147 wrote:
In relation to the sprint issue: what is strange to me is that Milano-San Remo was almost a flawless sprint result. Of the 11 favorites I listed in the preview, 9 favorites finished in the top-10. So it should be possible one way or another?
Unfortunately, PCM seems to make a difference between classics and stage races. Classics have been fine for most sprinters, but sprints in stage races for some reason behave quite differently - even if the field is just as stacked (or even more?).
Maybe the teams are stronger in general in the classics, might be interesting to e.g. run Qatar with the MSR startlist just to see whether that makes any kind of difference.
The main difference with PCM20 I can observe so far with flat classics is that late attackers don't seem to get as much love anymore. Personally, I'm fine with that, given that the flat beasts are often 50k riders who definitely can get their occasional success, but it was maybe a bit too much for my liking in the previous version. That's one plus I personally see with PCM22, but not everyone will agree 
|
| |
|
|
| SotD |
Posted on 13-11-2023 08:51
|
World Champion
Posts: 12187
Joined: 29-11-2006
PCM$: 2980.00
|
Also it could be interesting to split up a race like Qatar into 6 "one-day" races, and then calculate the time differences manually to see if the race would then make sense.
I suspect it could be the case. If yes, then we could atleast salvage quite a lot of the season for top sprinters, by manually calculating times in races that are obviously made for sprinters, and just leave the crap to races like Paris-Nice where the springers are playing second violin.
It requires a bit of knowledge prior to the season which races are ridden how, but it should be doable. Also with a bit of excel coding it should be somewhat easy to create a sheet where it calculates times for you if you just put in the bonifications from the intermediate sprints.
If someone would be willing to test Tour of Qatar like this, I would be very happy. I don't need anything but the results of each stage to setup the full results and see if it looks better, similar or worse.
|
| |
|
|
| TheManxMissile |
Posted on 13-11-2023 12:56
|
Tour de France Champion
Posts: 18187
Joined: 12-05-2012
PCM$: 0.00
|
Fabianski wrote:
Laurens147 wrote:
In relation to the sprint issue: what is strange to me is that Milano-San Remo was almost a flawless sprint result. Of the 11 favorites I listed in the preview, 9 favorites finished in the top-10. So it should be possible one way or another?
Unfortunately, PCM seems to make a difference between classics and stage races. Classics have been fine for most sprinters, but sprints in stage races for some reason behave quite differently - even if the field is just as stacked (or even more?).
Maybe the teams are stronger in general in the classics, might be interesting to e.g. run Qatar with the MSR startlist just to see whether that makes any kind of difference.
Stage Races vs Classics - a big part is what we explored with the GC Bug a year or two ago. In a Stage Race you have split Leader & Sprinter designations in the game, whilst in Classics you tend not to find that split, with the teams sprinter in a flat classic being designated as Leader. This does change the way the AI behaves, and I believe putting less emphasis on the "lead-out" and more on general positioning and protection.
Looking at how sprints are going currently this would in some part contribute to good classics AI. The team sprinters are not hampered by poor leadout AI, instead the team used not as a leadout but as a general pace setter, with leadout coming much later in proceedings and from different team riders. Basically the AI just works differently from Classics to Stage Races, based on if a sprinter is classed as Sprinter or Leader.
To tie it to SotDs point, if you ran Qatar as 6 classics I think you'd see this play out (and you could do the reverse with making an MSR stage race of 6 days of that), and have a better set of individual results. What you do loose however is the GC element, and chance of a shock breakaway win (i.e. what do we do if a breakaway does win one of those classics Qatar stages? Or how a surprise good result can put a rider into a GC spot where they & their team then works in a way to defend the GC).
Just to go further with the AI Leader/Sprinter thing, we also hit DB issues. In a more normal DB I think a good portion of teams would have a different rider designation in the game. Some teams would have a Leader set to a rouler/attacker, and then Sprinter set to their 2nd/3rd tier sprinter, changing how both the race overall works and sprints work. Our DB where most teams rock up with a 1st tier sprinter messes with this, seeing too many similar teams and therefore as the AI tries to sort out the designations you get a mess up where some riders get miss aligned and behave according to a different AI plan (i.e. as a Free Agent attacker or as a pure GC helper rather than working as we would expected a Sprinter to).
This is getting a bit long and messy. We're fighting to make the game do things it's not really designed to do. As making wholesale DB changes seems off the table, the idea of using multiple game versions so we can pick and choose the best things makes sense. But expand the conversation from sprinting, and establish what game does each element best and go all in on the idea.
|
| |
|
![imgur.com/KlJUqDx.jpg[/img[img]]https://imgur.com/yUygrQ.jpg](https://imgur.com/KlJUqDx.jpg[/img[img]]https://imgur.com/yUygrQ.jpg)
