Alle rapporten
De Crew rapporten
#25683: "BGA point given at the end of the game doesn't reflect the achievement"
Waarover gaat dit rapport?
Wat is er gebeurd? Selecteer uit het onderstaande
Suggestie: volgens mij zou het volgende de implementatie van dit spel enorm verbeteren
Gedetailleerde beschrijving
• Gelieve jouw suggestie precies en beknopt uit te leggen zodat het zo makkelijk mogelijk is om te begrijpen wat je bedoelt.
Hello,
Currently, whatever game was played, all players gains 1 and 1 point only.
Therefor, all players on BGA are still apprentice.
It is not possible to have the standard categories of "medium, good, expert" for this game for now.
I propose to give points according to the last succeed mission would help a lot.
Someone starting a campaign and stop playing at mission 2 would gain 2 points.
Someone starting at mission 20 and going to mission 35 would gain 35 points.
I personnaly don't want to start a mission at high level and getting people not used to the game at high difficulty at the table (and they can join by error not understanding it will be a hard one).
Another suggestion would be to have the sum of the mission succeeded as point, but I think it would go too high too quick.• Welke browser gebruik je?
Mozilla 78.3.0esr
Rapportgeschiedenis
3 okt 2020 10:09 • Jai le Leu • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
No other details
5 okt 2020 7:27 • ufm • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
Sum of the points per mission would work better because ratings stack. It doesn't have to be the same as the mission number... It can be instead like 'a tenth of the mission number per mission' or something.
5 okt 2020 13:42 • ufm • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
Also, the change of points at the game end will be useful to track games with higher/lower number of missions.
5 okt 2020 21:10 • Jai le Leu • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
It could also be the number of missions, with a negative count on failed tries.
Like mission succeed => +1
Failed attempt => -1
So it would does something like:
perfect => +1
2 tries => 0
3 tries => -1
etc...
This sumed for each mission would gives at the end a nice ratio.
I have witnessed a game that longed for more than 1h30, from mission 0 to mission 25... for sure the player at this table have a better experience than those who stop after mission 3 (and I don't want to play anymore with the laters...)
Like mission succeed => +1
Failed attempt => -1
So it would does something like:
perfect => +1
2 tries => 0
3 tries => -1
etc...
This sumed for each mission would gives at the end a nice ratio.
I have witnessed a game that longed for more than 1h30, from mission 0 to mission 25... for sure the player at this table have a better experience than those who stop after mission 3 (and I don't want to play anymore with the laters...)
7 okt 2020 5:20 • ufm • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
Flat number of cleared missions will not reflect the difficulty. At least it has to be incremental, if not exponential.
9 okt 2020 15:35 • ufm • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
Due to auto-matchmaking, players not used to the game join the tables for more difficult missions quite frequently. Player level classification by ELO is needed seriously.
9 okt 2020 16:18 • ufm • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
Also, some players join the game and quit immediately after 1 try to just raise their ELO quickly.
13 okt 2020 15:03 • matthiasvc • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
In order to have a proper rating system you'd need to have the system automatically award your rating based not only on how well you and your team did, but also on how this performance rates to other teams who played it. I would also assume your ELO-points might also depend on how many missions you did in a campaign.
So ordering my thoughts on requirements:
1) The better you do (win/lose ratio), the higher your ELO-rating should become!
2) Harder missions have a more difficult win/lose ratio!
3) Longer games should reward more scores, but in a linear way
4) The ELO-rating of other players shouldn't matter.
5) If you only win, your score needs to rise
The ELO rating BGA uses [K x (W - p(D))] has two weaknesses, the W says the result can only be Win-Loss-Tie whereas here we have a string of victories and losses. And the p(D) requires the rating of an adversary where there is none in the game. the K can be used as elasticity factor, but the length of the game should be also taken into account for this elasticity.
I was thinking about a quite elaborate set-up for how to overcome these problems, but it might be a lot more simpler. Why not treat every mission as a player whom you're all playing against? If you lose, the mission's rating goes up by a tiny bit (probably allow for an ever diminishing elasticity). If you win it's as if you win against a player. This would mean your ranking would be evaluated after every single mission and not only after a campaign solving any or all issues we have currently.
So ordering my thoughts on requirements:
1) The better you do (win/lose ratio), the higher your ELO-rating should become!
2) Harder missions have a more difficult win/lose ratio!
3) Longer games should reward more scores, but in a linear way
4) The ELO-rating of other players shouldn't matter.
5) If you only win, your score needs to rise
The ELO rating BGA uses [K x (W - p(D))] has two weaknesses, the W says the result can only be Win-Loss-Tie whereas here we have a string of victories and losses. And the p(D) requires the rating of an adversary where there is none in the game. the K can be used as elasticity factor, but the length of the game should be also taken into account for this elasticity.
I was thinking about a quite elaborate set-up for how to overcome these problems, but it might be a lot more simpler. Why not treat every mission as a player whom you're all playing against? If you lose, the mission's rating goes up by a tiny bit (probably allow for an ever diminishing elasticity). If you win it's as if you win against a player. This would mean your ranking would be evaluated after every single mission and not only after a campaign solving any or all issues we have currently.
13 okt 2020 18:05 • ufm • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
Current Hanabi rating works like that: forum.boardgamearena.com/viewtopic.php?t=4814
But can we assign ratings for all 50 individual missions? That might be overcomplicated.
Thermopyles and Bandido rating works well without ELO system (these game adds/subtracts flat amount based on the result)
But can we assign ratings for all 50 individual missions? That might be overcomplicated.
Thermopyles and Bandido rating works well without ELO system (these game adds/subtracts flat amount based on the result)
15 okt 2020 23:06 • Jai le Leu • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
I like matthias idea of considering each mission as a player ;)
21 nov 2020 23:59 • diart • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
You gain a point even if you lose a mission. Now people are playing one round at a time.... so as to gain an ELO point for each round played ... individually. While I don't want to be dinged badly for every round a team loses....having the ELO simply based on how many rounds/games you played seems a little..? not reflective of skill?
10 dec 2020 22:09 • Supergravity • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
I was thinking of how to improve the rating system in this, and I really like the matthias idea better than anything I have been able to come up with. However, it might clash with some overall site framework for how ratings are adjusted.
If the framework demands just one result agains an opponent, a way around that would be to compute a single rating for a hypothetical opponent, which starts at 100 and goes up (by some amount that depends linearly on the mission number) for each success and down (by some fixed amount) for each failure. Then at the end, everybody ties that hypothetical opponent.
If the framework demands just one result agains an opponent, a way around that would be to compute a single rating for a hypothetical opponent, which starts at 100 and goes up (by some amount that depends linearly on the mission number) for each success and down (by some fixed amount) for each failure. Then at the end, everybody ties that hypothetical opponent.
13 dec 2020 22:40 • Vynce • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
Matthias' suggestion is probably the cleanest.
I was going to suggest basing it on (total value of missions completed - fails) / (attempts) where each mission was worth a number of points equal to its mission number, but then you'd still need to measure it against something. Matthias' is self-balancing.
I was going to suggest basing it on (total value of missions completed - fails) / (attempts) where each mission was worth a number of points equal to its mission number, but then you'd still need to measure it against something. Matthias' is self-balancing.
17 mrt 2021 22:10 • Tisaac • De ontwikkelaars willen graag meer informatie over deze suggestie:
Please see and go the forum thread for an explanation of what is actually possible.
18 mrt 2021 9:43 • recontact • De ontwikkelaars willen graag meer informatie over deze suggestie:
@Tisaac : could you provide us the link to the Forum thread 🙏🏻 ?
18 mrt 2021 9:56 • Tisaac • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
Here it is :
boardgamearena.com/forum/viewtopic.php?f=293&t=19261
boardgamearena.com/forum/viewtopic.php?f=293&t=19261
1 okt 2022 14:35 • yfimc • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
please add proper elo for crew and deep sea crew. it is so annoying to try playing cooperative games with people whose skill level doesn’t match their experience. it’s dismaying there is still no activity on this suggestion after years of requests
11 feb 2023 13:05 • ufm • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
2 mei 2023 10:44 • Mioranduh • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
If you spent two hours playing through several missions you only get ONE single ELO point. But if you spend just 5 minutes to finish the first mission you get ONE ELO point too. That doesn't make any sense!
Also the statistics do not even show which missions you have played and how many times you have tried to do them. But while playing offline this is how we are supposed to play it according to the rulebook. Writing down in how many attempts you achieved the missions. But the statistics only show the result (1 ELO) and the thinking time. Compared to nearly all BGA games, this is very limited.
Also similar to the report in the other Crew game:
#80070: "Award 1 ELO per mission completed"
boardgamearena.com/table?table=340947946
Also the statistics do not even show which missions you have played and how many times you have tried to do them. But while playing offline this is how we are supposed to play it according to the rulebook. Writing down in how many attempts you achieved the missions. But the statistics only show the result (1 ELO) and the thinking time. Compared to nearly all BGA games, this is very limited.
Also similar to the report in the other Crew game:
#80070: "Award 1 ELO per mission completed"
boardgamearena.com/table?table=340947946
10 jul 2024 8:21 • hanabifanabi • De ontwikkelaars hebben deze suggestie nog niet geanalyseerd:
I would also prefer to have 1 point per mission completion, to most accurately represent a person's experience with the game. This would also solve the issue of people farming mission 10 repeatedly for "ELO".
Dit rapport aanvullen
Gelieve hier alle informatie toe te voegen die relevant is om deze bug te reproduceren of jouw suggestie te begrijpen:
- Andere tafel ID / zet ID
- Loste F5 het probleem op?
- Kwam het probleem meerdere keren voor? Altijd? Willekeurig?
- Als je een schermafbeelding van deze bug hebt (goede gewoonte), kun je Imgur.com gebruiken om deze te uploaden, de link kopiëren en hier plakken.
