Update 44 is now available for testing on the PTS! You can read the latest patch notes here: https://forums.elderscrollsonline.com/en/categories/pts
Maintenance for the week of October 7:
• PC/Mac: No maintenance – October 7
• Xbox: EU megaserver for maintenance – October 9, 2:00 UTC (October 8, 10:00PM EDT) - 16:00 UTC (12:00PM EDT)
• PC/Mac: No maintenance – October 7
• Xbox: EU megaserver for maintenance – October 9, 2:00 UTC (October 8, 10:00PM EDT) - 16:00 UTC (12:00PM EDT)
Broken Leaderboard Needs Overhauling
RCubed1967
✭✭
Tales of Tribute is one of the best strategy games I have played. I thoroughly enjoy it, but absolutely hate the ranking system. Due to this, I decided to evaluate why and see if I could offer improvements.
In April 2023, I tracked each of my 150 ranked games played: who I played, who had first pick, my score after each match, and special notes such as patron wins. After each game, I checked the leaderboards for my current score before joining the queue for another game.
A graphic of my scores is shown at the end. I ended up in the +10% which is not my best. I travel overseas a lot, so my game play varies. You can see my jet lag losses after my 11-game winning streak (note to self – don’t play tired lol).
The results show a significant need for overhaul. I will first discuss my major issues and observations, and finally my recommendations for changes.
Major Issues and observations
1. The score for a win or loss seems to be affected by a max point (difference in rank between players) and diminishing returns for subsequent wins or losses.
1.1. The early season is substantially affected by diminishing returns.
1.1.1. Opens the door to exploitation as a player can choose to lose after 3 or 4 wins to reset diminishing returns.
1.2. Diminishing returns is less relevant in the later game.
1.2.1. Many more players are being ranked.
1.3. The methodology is not perfect as I won three consecutive games against one player that did not meet the max point and diminishing returns (and I was coming off a loss).
1.3.1. Score gains of +34, +29, then +31
1.4. Massive swings in scores were occurring.
1.4.1. My 11-game winning streak in the early season gained me 684 points while a late-season 5-game winning streak gained me 622 points.
2. I had two wins with ZERO score gains.
2.1. I was only playing to avoid losing points!
3. Once I played a person three times in a row, I won the first two games and lost the third...yet I lost points (+70, + 47, -150 = net -33).
3.1. To be ranked means that winning a series should have positive gains!
4. I had one game that the computer failed to load me in, and I lost 79 points plus I got a deserter penalty!
4.1. This was reported as a bug.
4.1.1. Notably, I remember several fail-to-load occurrences in March.
5. I played against others who forfeited and played me immediately (seconds) after
5.1. A cheat exists to bypass the deserter penalty.
5.1.1. This was reported as a bug and an exploit.
6. The first player advantage is still too big.
6.1. I had 74 starting picks and 75 second picks (plus the failed load loss). With the first pick, I won 59% of the games versus 48% without it.
6.1.1. One loss was due to a computer crash. I accept that as a loss (although I was winning).
7. It appears several individuals stop playing after a certain point to hold their score.
7.1. This is an exploitation of the game for two reasons:
7.1.1. Very tough to overtake a player if they don’t play.
7.1.2. Players could switch to second accounts and play to double down rewards.
7.1.2.1. A golfer should not get double rewards for playing the same tournament with two different golf bags.
7.1.2.2. Ranking is about the player, not the number of accounts they have access to!
8. The swings are too big (especially in the late game).
8.1. I had back-to-back losses (different players) of -150 each.
8.2. I had back-to-back wins against the same player for a gain of 150 each.
8.3. When the top 10% has players under 1000 points, a 300-point swing in 2 games is ridiculous!
9. The activity finder rank is not up to date between games.
9.1. Players need to check the leaderboard to update the current rank.
10. I played Druid Braigh for the last druid card (won first game) and dropped 5 points in score.
10.1. A win against an NPC affects leaderboards (reported as a bug)? Happened end of month.
Recommendations
11. Eliminate the need to get to Rubedite level for seasoned players.
11.1. When a player has played 10 games in a month at the Rubedite level, automatically enter them at the Rubedite level in the next month.
11.1.1. The games played to reach Rubedite are better served in the ranking system.
12. Normalize wins and focus on sets between players.
12.1. A player enters ranked play with 500 points which is modified with time of entry to recognize consistent players.
12.1.1. Subtract 10 points per day into the month.
12.1.1.1. A player entering ranked play on day 5 starts with 450 points.
12.1.1.2. A player entering ranked play on day 25 starts with 250 points.
12.2. A win is worth 5 points, a loss is worth -5 points.
12.2.1. The type of win or the rank of player that you are playing against do not matter.
12.3. Series against another player over the month is worth 40 points.
12.3.1. When two players have played and one has won 2 games, the winner of two games is awarded all 40 points.
12.3.1.1. Logic formula between 2 players (table shown at bottom)
12.3.1.1.1. If wins by either player against each other>1
12.3.1.1.1.1. 40 * wins/(wins + losses)
12.3.1.1.2. Round to 0 digits
12.3.2. As the players play more games against each other over the month, the 40 points is split between the two players.
12.3.2.1. Split is based on (40 * ((wins – 1)/(total games)))
12.3.2.1.1. No matter how many games the two players play against each other, the total points distributed between the two players for the series is 40 points.
12.3.2.1.2. The +5/-5 points for win/loss is still added on top of the series value.
12.4. Set the queue to prioritize matches between people who do not have series points distributed between them.
12.4.1. Keep players in queue for up to 30 seconds extra to help facilitate this.
12.5. Players where one player has 4 wins against another player cannot play against that player again for the remainder of the month.
12.6. The focus on games and series will allow for more realistic ranking.
12.6.1. Players cannot sit back on a high point total thinking no one will catch them.
12.6.1.1. Dissuades switching to second account.
12.6.2. Massive score swings are eliminated.
12.7. Note that hat I only played against 3 of the top 10 all month. The first had me 2-2 (would have been 3-1 if my computer hadn’t crashed), I lost 2-1 against a second player, and in the third I lost my only game against them.
13. Require that a game is only counted after the first player ends their turn.
13.1. Failure to load would not be an issue.
14. Remove the deserter penalty.
14.1. Allow players to lose graciously without preventing them from playing again immediately!
15. Reevaluate the cost of early game power cards.
15.1. There are some cards so valuable at the start that the second pick cannot catch up.
15.1.1. These cards may not be captured in the first pick but will be in the third pick (after the first shuffle).
16. Introduce a mobile game, linked to our ESO account, so we can play without being on our computer.
16.1. No extra cost to ESO Plus players
16.1.1. Allows more flexibility for playing games.
16.1.2 Ding, Ding, Ding - opportunity for ZOS to make money from non-ESO Plus members AND non-ESO player too
17. Add current score to the group finder and current rank to the activity finder.
17.1. Consider adding a current series update on game start.
17.1.1. E.g., Series Game 5 vs. Player A. Player A has 2 wins, and you have 2 wins.
18. Penalize players for playing multiple accounts in the same tournament.
18.1. There are a few ways to check for this.
Point Distribution Table
My April Scores
In April 2023, I tracked each of my 150 ranked games played: who I played, who had first pick, my score after each match, and special notes such as patron wins. After each game, I checked the leaderboards for my current score before joining the queue for another game.
A graphic of my scores is shown at the end. I ended up in the +10% which is not my best. I travel overseas a lot, so my game play varies. You can see my jet lag losses after my 11-game winning streak (note to self – don’t play tired lol).
The results show a significant need for overhaul. I will first discuss my major issues and observations, and finally my recommendations for changes.
Major Issues and observations
1. The score for a win or loss seems to be affected by a max point (difference in rank between players) and diminishing returns for subsequent wins or losses.
1.1. The early season is substantially affected by diminishing returns.
1.1.1. Opens the door to exploitation as a player can choose to lose after 3 or 4 wins to reset diminishing returns.
1.2. Diminishing returns is less relevant in the later game.
1.2.1. Many more players are being ranked.
1.3. The methodology is not perfect as I won three consecutive games against one player that did not meet the max point and diminishing returns (and I was coming off a loss).
1.3.1. Score gains of +34, +29, then +31
1.4. Massive swings in scores were occurring.
1.4.1. My 11-game winning streak in the early season gained me 684 points while a late-season 5-game winning streak gained me 622 points.
2. I had two wins with ZERO score gains.
2.1. I was only playing to avoid losing points!
3. Once I played a person three times in a row, I won the first two games and lost the third...yet I lost points (+70, + 47, -150 = net -33).
3.1. To be ranked means that winning a series should have positive gains!
4. I had one game that the computer failed to load me in, and I lost 79 points plus I got a deserter penalty!
4.1. This was reported as a bug.
4.1.1. Notably, I remember several fail-to-load occurrences in March.
5. I played against others who forfeited and played me immediately (seconds) after
5.1. A cheat exists to bypass the deserter penalty.
5.1.1. This was reported as a bug and an exploit.
6. The first player advantage is still too big.
6.1. I had 74 starting picks and 75 second picks (plus the failed load loss). With the first pick, I won 59% of the games versus 48% without it.
6.1.1. One loss was due to a computer crash. I accept that as a loss (although I was winning).
7. It appears several individuals stop playing after a certain point to hold their score.
7.1. This is an exploitation of the game for two reasons:
7.1.1. Very tough to overtake a player if they don’t play.
7.1.2. Players could switch to second accounts and play to double down rewards.
7.1.2.1. A golfer should not get double rewards for playing the same tournament with two different golf bags.
7.1.2.2. Ranking is about the player, not the number of accounts they have access to!
8. The swings are too big (especially in the late game).
8.1. I had back-to-back losses (different players) of -150 each.
8.2. I had back-to-back wins against the same player for a gain of 150 each.
8.3. When the top 10% has players under 1000 points, a 300-point swing in 2 games is ridiculous!
9. The activity finder rank is not up to date between games.
9.1. Players need to check the leaderboard to update the current rank.
10. I played Druid Braigh for the last druid card (won first game) and dropped 5 points in score.
10.1. A win against an NPC affects leaderboards (reported as a bug)? Happened end of month.
Recommendations
11. Eliminate the need to get to Rubedite level for seasoned players.
11.1. When a player has played 10 games in a month at the Rubedite level, automatically enter them at the Rubedite level in the next month.
11.1.1. The games played to reach Rubedite are better served in the ranking system.
12. Normalize wins and focus on sets between players.
12.1. A player enters ranked play with 500 points which is modified with time of entry to recognize consistent players.
12.1.1. Subtract 10 points per day into the month.
12.1.1.1. A player entering ranked play on day 5 starts with 450 points.
12.1.1.2. A player entering ranked play on day 25 starts with 250 points.
12.2. A win is worth 5 points, a loss is worth -5 points.
12.2.1. The type of win or the rank of player that you are playing against do not matter.
12.3. Series against another player over the month is worth 40 points.
12.3.1. When two players have played and one has won 2 games, the winner of two games is awarded all 40 points.
12.3.1.1. Logic formula between 2 players (table shown at bottom)
12.3.1.1.1. If wins by either player against each other>1
12.3.1.1.1.1. 40 * wins/(wins + losses)
12.3.1.1.2. Round to 0 digits
12.3.2. As the players play more games against each other over the month, the 40 points is split between the two players.
12.3.2.1. Split is based on (40 * ((wins – 1)/(total games)))
12.3.2.1.1. No matter how many games the two players play against each other, the total points distributed between the two players for the series is 40 points.
12.3.2.1.2. The +5/-5 points for win/loss is still added on top of the series value.
12.4. Set the queue to prioritize matches between people who do not have series points distributed between them.
12.4.1. Keep players in queue for up to 30 seconds extra to help facilitate this.
12.5. Players where one player has 4 wins against another player cannot play against that player again for the remainder of the month.
12.6. The focus on games and series will allow for more realistic ranking.
12.6.1. Players cannot sit back on a high point total thinking no one will catch them.
12.6.1.1. Dissuades switching to second account.
12.6.2. Massive score swings are eliminated.
12.7. Note that hat I only played against 3 of the top 10 all month. The first had me 2-2 (would have been 3-1 if my computer hadn’t crashed), I lost 2-1 against a second player, and in the third I lost my only game against them.
13. Require that a game is only counted after the first player ends their turn.
13.1. Failure to load would not be an issue.
14. Remove the deserter penalty.
14.1. Allow players to lose graciously without preventing them from playing again immediately!
15. Reevaluate the cost of early game power cards.
15.1. There are some cards so valuable at the start that the second pick cannot catch up.
15.1.1. These cards may not be captured in the first pick but will be in the third pick (after the first shuffle).
16. Introduce a mobile game, linked to our ESO account, so we can play without being on our computer.
16.1. No extra cost to ESO Plus players
16.1.1. Allows more flexibility for playing games.
16.1.2 Ding, Ding, Ding - opportunity for ZOS to make money from non-ESO Plus members AND non-ESO player too
17. Add current score to the group finder and current rank to the activity finder.
17.1. Consider adding a current series update on game start.
17.1.1. E.g., Series Game 5 vs. Player A. Player A has 2 wins, and you have 2 wins.
18. Penalize players for playing multiple accounts in the same tournament.
18.1. There are a few ways to check for this.
Point Distribution Table
My April Scores
0
-
spartaxoxo✭✭✭✭✭
✭✭✭✭✭RCubed1967 wrote: »5. I played against others who forfeited and played me immediately (seconds) after
5.1. A cheat exists to bypass the deserter penalty.
5.1.1. This was reported as a bug and an exploit.
After a certain amount of time, there is no deserter penalty. I think it's like 5 to 10 minutes, somewhere in there. In addition, sometimes the deserter penalty just randomly doesn't apply. The player has no control over it afaik.0 -
NeKryXe✭✭✭✭✭
✭spartaxoxo wrote: »RCubed1967 wrote: »5. I played against others who forfeited and played me immediately (seconds) after
5.1. A cheat exists to bypass the deserter penalty.
5.1.1. This was reported as a bug and an exploit.
After a certain amount of time, there is no deserter penalty. I think it's like 5 to 10 minutes, somewhere in there. In addition, sometimes the deserter penalty just randomly doesn't apply. The player has no control over it afaik.
I've been noticing that lately. It's better than nothing, but maybe on ranked it's not needed, because it's very rare for someone to just quit for nothing on a ranked match.
But on casual it's a completely different story. Many players, really a lot, quits very fast, sometimes even in the first hand if they don't get something to increase their chances. Even with time penalty I met a lot of players clearly trying their luck and quitting, so if there was no penalty probably all matches would be one hand and the one who gets the worst hand quits and immediately joins a new match.
I don't think we need any penalty on ranked matches, but we clearly need on casual.1 -
DragonRacer✭✭✭✭✭
✭✭If a player deserts on their Character A, they can simply switch to Character B and rejoin the queue with no deserter penalty, so I have been told. So long as those characters have done the starting ToT mission to be able to join queues, a player could use any of them freely like that.
This is probably what you encountered. It's not like we get to see our opponent's character in order to see "Hey, you were a Khajiit 2 minutes ago when you quit, now you are an Argonian!"PS5 NA. GM of The PTK's - a free trading guild (CP 500+). Also a werewolf, bites are free when they're available. PSN = DragonRacer131 -
AnduinTryggva✭✭✭✭✭
✭Re-queuing with another character is not exploiting as it is not a bug. Punishable exploiting only applies to bugs.
I think as long as they have the futile time penalty upon concede it is ok to relog with another character as long as it is the same account.
For the ranking system I'd add a daily default point loss in order to prevent players sitting on their score.1 -
sekou_trayvond✭✭✭✭✭OP, this is single-handely the best, most informative addition to the TOT forum to date. Srsly. Well done. Cold hearted data is much appreciated.
Very much approve of your Point 11 recommendations. As for Point 12.1, I understand what you're trying to do here but I'd suggest that it penalizes people who come in to the latest season a little later for no other reason than the vagaries of life affecting one's playtime schedule.
I will add more thoughts as I look over it some more.
Edited by sekou_trayvond on May 8, 2023 1:30AM1 -
Rooatouille✭✭As @sekou_trayvond said, thank you for sharing this data -- it is very much appreciated. Again, like Sekou, I could agree with your suggestion in section 11, but I think your section 12 suggestion unnecessarily overcomplicates things. The ranked point gain and loss absolutely needs changes and I can agree with the need for a decay system (i.e., once you reach a certain rank you start to lose X RP/day if you don't play at least one game). How I've suggested fixing the RP distribution before and how I still think it would work best is to add a base RP gain for wins with an MMR-adjusted pool of RP on top of that to a maximum of 150 RP per win. I suggest a base 60 RP gain.
I logged in tonight and started at rank 3 on the leaderboard. After playing 5 games with 1 loss for -150, I'm sitting at a net -40 RP because the gains for winning those 4 games is just too low. I also had a number of games I won yesterday that were the classic +0 RP gained. I don't believe one should receive the same number of points for a win as they lose for a loss without an MMR adjustment, but those +0 games feel so bad. A player's MMR should be a consideration -- that's how you come out with a leaderboard that's representative of player skill rather than raw quantity of games. MMR systems are also a tried and true method used in a variety of other ranking systems, so I don't think we need to reinvent the wheel here, we just need to adjust the ranked system to better suit ToT's needs, which they've done before (the days of -1000 RP are thankfully no longer with us lol).
All in all, fantastic data given and thank you for providing your thoughts, the community absolutely needs members like you willing to pay attention and provide feedback!1