2017 PBEM Duel Tournament – stats and results

Home Forums Age of Wonders 3 Discussions 2017 PBEM Duel Tournament – stats and results

This topic contains 11 replies, has 7 voices, and was last updated by  marcuspers1 3 months ago.

Viewing 12 posts - 1 through 12 (of 12 total)
  • Author
    Posts
  • #254703

    Hiliadan
    Member

    The 2017 PBEM Duel Tournament is now on its 3rd round. Players cannot replay a race or a class they played in previous rounds so they have to play all 7 classes and they need to pick 7 out of the 9 races during their first 7 matches (they can start replaying the class and race again on round 8 if they went that far!). It means players have to choose the best race for each class, while thinking about other classes (e.g. Dwarf Theocrat is very good, but Dwarf Dreadnought too, so should I take Dwarf for Theo or Dread?). As you also get eliminated if you lose (twice, as it’s double elimination), it also means you need to play a combo strong enough to beat the opponents you’re facing each round. So you have both incentives to keep your best combos for the later stages (you clearly want your best combo for round 7 which should be among the most difficult, as your opponent will also have beaten all or all but 1 of his opponents and be very good!) and to avoid using a very weak combo if your opponent in an early round is very strong.
    All in all, the classes and races chosen during the first 3 rounds probably do not contain the best combos (e.g. few Necros) because the best players are keeping them in their sleeves for later rounds, but it still provide good insights on what are the best combinations of races and classes. So let’s have a look at the key figures from the stats: https://www.the-battlefield.com/aow3/index.php?page=ladderstats&tourneyid=6

    1/ All classes are more or less equally represented except Arch Druid (8% vs an expected 14% if all classes were equally represented) and Necro (9%). This is rather surprising for AD. For Necro, it probably indicates the top players are keeping it for later rounds. Among the winner bracket, DarkRider, Olop and Ezekiel have already used their Necro combos while gabthegab, Jean de Metz and rrrrookie haven’t.

    2/ Races are not equally represented at all (they would all be chosen at 11% if they were): Dwarves (17%), Goblins (16%) and Orcs (17%) stand out, probably because they are easier races to play and weaker players chose them. Dwarves and Goblins indeed have very poor results, with a victory ratio of only 25%. The top players will have to play 7 different races in any case, which means each will not play 2 races, which ones could it be? Several races are currently not very liked: 3 are at 7% (Elves, Tigrans, Humans) while Halflings do a little better at 8%. Tigrans and Humans are probably kept by several for later rounds, so the two “unplayed” races might be Elves and Halflings.

    3/ Warlord appears to be the weakest class for now, with only 2 victories out of 13 matches (only Jean de Metz and gabthegab, the two top players, managed to pull it off)!
    Dreadnought also struggles with only 2 out of 9 victories.

    4/ Necro does well with 80% victories but we have only 5 matches to look at (the last one was a mirror match it seems).

    5/ Tigrans do well too (80%) but here too, the number of matches is limited to 5.

    6/ In terms of favourite race/class combinations, we have the following:
    AD: no favourite race, 4 races are not (yet) played: Dwarves, Elves, Frostlings, Humans, but that’s probably because the number of matches is limited (only 6 AD so not all races could be represented anyway)
    Dread: Dwarves dominate with 7 Dwarf Dread vs only 1 or 2 for other races. Surprisingly, Human Dread is not played, probably because Human is better combined with other classes.
    Necro: no favourite race. One of the favourite Necro, Draconian, is not represented. Probably because several top players have not yet played Necro, and also because Draconian is better combined with Sorcerer (5 out of the 8 Draconian chose Sorcerer).
    Rogue: Goblin Rogue arrives on top (4 out of 13), then Orcs (3), Human (2) and Tigran (2). 6 races are represented.
    Sorcerer: domination of Draconian Sorcerer (5 out of 10), with its top Apprentice and Shrine cost reduction in Racial Governance. Honourable mention for Frostlings (3 out of 10).
    Theo: 7 races represented, Dwarves and Orcs are the only one with more than 1 example, with 4 players choosing them each.
    Warlord: 7 races too, with Orcs on top (4 out of 13) but no success at all (100% defeat, vs Rogue Frostling, AD Tigran, Sorcerer Elf, and Theocrat Human). The distribution of Orc races probably reflect more the choices of players not very used to the class, and relatively weaker players, as well as a willingness to play the weakest race and combos in the first rounds, to keep the better ones for later rounds.

    Let’s see how the stats evolve as we move toward later stages of the tournaments!
    Welcome to discuss my analysis and the results! 🙂

    • This topic was modified 1 year ago by  Hiliadan.
    • This topic was modified 1 year ago by  Hiliadan.

    The AoW3 Wiki, including a video library!
    Improve AoW3 with the balance mod (presentation video)
    The Battlefield.com/AoW3 the community website for AoW3

    #254808

    AlXStormrage
    Member

    Hillia, there are 3 Warlord wins out of 14 games (not sure if you were looking at statistics before the last reported game) 🙂

    Also I believe the top 3 classes to dominate rankings are:
    Theocrat (71% win-ratio)
    Necro (67% win-ratio)
    Sorcerer (64% win-ratio)

    which, in my opinion, is the close representation of the top classes strength (perks, auras, bonuses, spells). Also, I don’t know if anyone has made any analysis of this, but from my experience, most of the players are trying to rush their opponent in the duel, i.e. players tend to do some site clearing along the way and to raid opponent’s cities asap. Which is why classes that have strong start are getting better results. The strategic layer and long-term decisions are of low importance, as most game tend to end at around round 30.

    I do believe that the Warlord is underrated, as the potential of the class is pretty high, but it has weaker scouting capabilities (scouts must be built, taking precious production-turns), comparing to the top 3.
    I believe that Archdruid and Rogue are well-balanced at the moment, but due to AD now requiring harder grinding (for evolution of the spiders and baby serpents) as well as BA skill available at later stages – rush attacks are not well suited for it at the moment. Rogue, IMHO, is also stronger at the mid-game, with it’s underhanded spells and stealth tactics.
    As for the Dreadnought – it’s also class that can show best power at the mid-late stages of the game, although it can do site clearing pretty well at the initial stages.

    #254811

    AlXStormrage
    Member

    Scheming during the work 🙂

    Attachments:
    You must be logged in to view attached files.
    #254816

    Hiliadan
    Member

    Yes, the stats were produced before your win as Warlord (congratz! but that confirms the trend so far, only top 4 players have been able to win with Warlord so far) and the defeat of Gladis as Necro who surrendered his game early as he is kind of leaving AoW3, which somehow blurred and degraded the stats for Necro.

    I’m not so sure Sorcerer is in the top 3 class, I would say Theo, Rogue or AD are all superior to Sorcerer on our settings, and especially our map size. Sorcerer has no mind control and even if evolving is more difficult, you can evolve a T4 by turn 30 when you will need it to end the game (as I did vs chiveicrook for instance).
    Sorcerer’s scouting is also worse than Theo or Rogue (the 2 best) IMO.

    Also, I don’t know if anyone has made any analysis of this, but from my experience, most of the players are trying to rush their opponent in the duel, i.e. players tend to do some site clearing along the way and to raid opponent’s cities asap. Which is why classes that have strong start are getting better results. The strategic layer and long-term decisions are of low importance, as most game tend to end at around round 30.

    Yes, I agree and that’s something we still need to work on. It’s clear that smaller maps will always mean a lesser role for strategy and a bigger role for clearing. But I believe that for instance removing the +50% reward from Strong Defenders and adding battle enchantment to all Legendary and Mystical sites that lack one will help to slow things down and make clearing less important, increasing the strategic depth.

    Warlord’s biggest potential in my eyes is that after a few level and with the right spells, it’s actually the strongest clearing class. Is that also what you mean when you say it is under-rated?

    The AoW3 Wiki, including a video library!
    Improve AoW3 with the balance mod (presentation video)
    The Battlefield.com/AoW3 the community website for AoW3

    #254819

    AlXStormrage
    Member

    Warlord’s biggest potential in my eyes is that after a few level and with the right spells, it’s actually the strongest clearing class. Is that also what you mean when you say it is under-rated?

    Partially so. Warlord clears well, but IMO the Dreadnought has just as good clearing options.
    WL main strength lies in the fact, that with the right skill set, WL units will cost almost nothing. For example in my game I had bought berserkers for less than 50 gold (43?) and something like 9 mana (I am trying to remember now, but not sure if those were accurate prices), which is ridiculous if you compare the strength/price.

    I’m not so sure Sorcerer is in the top 3 class, I would say Theo, Rogue or AD are all superior to Sorcerer on our settings, and especially our map size. Sorcerer has no mind control and even if evolving is more difficult, you can evolve a T4 by turn 30 when you will need it to end the game (as I did vs chiveicrook for instance).
    Sorcerer’s scouting is also worse than Theo or Rogue (the 2 best) IMO.

    Sorcerer doesn’t have mind control – true, but it usually gets at least one hero that can charm/convert/befriend. Sorcerers main strength lie in the ability to clear heavy sites with the help of PW, floating units and also t2 units that can stun (you can get 1-2 stacks pretty fast and if SoE researched you get stun for them). Also don’t forget – at higher level Sorcerer can take MoI, which renders army invisible.
    As for AD – I never said it’s bad 🙂 on the contrary – I believe AD is well balanced now, just the majority of the players don’t find it as easy to play, as before the patch, when units evolved like crazy

    #254849

    xlnt
    Member

    Heyya,

    i’m with ALXSTORMRAGE on this one – there are clear T1 combos and then there are the rest and IMO there are the T3 ones too – but i accept that this is arguable.

    T1: is definitely Theo and i will put Necro and Rogue there too – but if i have to win, i’ll go Theo.

    T2: Sorc might be the best T2 – but it’s not T1 imo – too dependent on t2 summons and for a lot of time, also research buildings… i believe the AD is almost the same.

    T3: WL and DN – those can win vs T2 but require a lot of effort vs T1 combos IMO. You have to be better and you have to do something else rather then cross the map while getting a stack of T3 units on the way (: WL can be frustrating with currents ON the map ): and DN … what can i say – it’s not for everyone. it may not be THAT bad but it surely is not fun or easy to understand/use or clear to see the strategies open for the player. in short – it’s a pain ):

    • This reply was modified 1 year ago by  xlnt.
    #254863

    swups
    Member

    can someone post the map settings and some general rules fot this tournament,
    like using mod or not…

    i only got thats a small map.

    thx

    #254865

    Hiliadan
    Member

    It’s a Tournament size map (57×49) actually, not a Small map (49×33).
    You can see the full settings and rules here: https://www.the-battlefield.com/aow3/index.php?page=rules&tourneyid=6
    So it uses the PBEM & Single Player balance mod: http://steamcommunity.com/sharedfiles/filedetails/?id=661597466

    The AoW3 Wiki, including a video library!
    Improve AoW3 with the balance mod (presentation video)
    The Battlefield.com/AoW3 the community website for AoW3

    #271148

    Hiliadan
    Member

    We are now on Round 8 of the Tournament, with only 3 matches left, and all but 2 players played all their available combos so the stats are now much more complete!
    I made a preliminary analysis based on the meta stats: https://docs.google.com/spreadsheets/d/e/2PACX-1vRf3H70tDsrj6cqo-gQn-E0g8QFWaDjFQiwVZ79ORNYVgXSG4TEGQRUsJJrO4JZXu_9R0jKSs8fiOI4/pubhtml?gid=1201203921&single=true

    So let’s update the analysis from July 2017!

    1/ All classes are more or less equally represented
    This is obviously still true due to the rules of the tournament, which force players to play at least once each class. However, we can see that Rogue, Theocrat, Warlord and Sorcerer got played more than the rest. That is not due to the best players choosing them for their 2nd choice because such “2nd play of class” will happen only after round 8 (and it’s still ongoing so its stats are not yet counted, though the ongoing match opposes an Arch Druid to a Warlord). Rather, it seems to be that most players eliminated in early rounds played these 4 classes and that top players (Ezekiel, rrrrookie, Hiliadan, marcuspers, Jean de Metz, gabthegab, who got up to round 7) also played them, except:
    – Ezekiel didn’t play Rogue, probably keeping it for the final matches (but he got eliminated)
    – rrrrookie didn’t play Warlord
    – gabthegab didn’t play Theocrat and Rogue, keeping them for his 2 last matches, probably because he judges them among the top 2 classes

    2/ Races are not equally represented at all (they would all be chosen at 11% if they were)
    When round 3 was still ongoing, Dwarves (17%), Goblins (16%) and Orcs (17%) stood out, and I proposed this explanation “probably because they are easier races to play and weaker players chose them” and noted that “Dwarves and Goblins indeed have very poor results, with a victory ratio of only 25%”. I guessed that “the two “unplayed” races might be Elves and Halflings”.
    Dwarves (14%), Goblins (13%) and Orcs (15%) remain the top 3 race choices and both Dwarves and Goblins keep very bad results (25% and 33% respectively) but Orcs are one of the most successful race with a win-ratio of 59%. Jean de Metz picked Orcs and won against gabthegab (his first defeat ever in ranked PBEM!!) on round 7 while Ezekiel picked Goblin and lost vs Marcus on the same round. So it seems Orcs are not just an easy race but also a strong one!
    Conversely, Frostlings appear relatively weak, with few picks and average (50%) win ratio.

    3/ Warlord still appear to be one of the weakest class but Dreadnought is barely better
    On round 3, Warlord had only 2 victories out of 13 matches (only Jean de Metz and gabthegab, the two top players, managed to pull it off)! At the end of round 7, it’s barely better, with 4 victories, the 2 additional winners being Hiliadan and AlXStormrage (top 5 player in the overall ranking who surrendered early in his round 6 match vs marcuspers).
    Dreadnought struggled with only 2 out of 9 victories and also barely improved that with 4 out of 14 victories.
    The round 8 match between a Human Arch Druid and a Tigran Warlord (marcus vs Hiliadan) will show if Warlord can win against veteran opponents, or if, like Dreadnought, it struggles too much due to its inner weaknesses. As AlXStormrage pointed out in this thread, Warlord is probably under-rated, but struggles with scouting.

    4/ Theocrat replaces Necro as the top class, with 75% victories (15 out of 20)

    5/ Tigrans crash completely, from 80% to 33%, losing all their matches after round 3.
    gabthegab lost as Tigran Necromancer vs Jean’s Orc Sorcerer, and, notably, Jean hasn’t played his Tigran pick yet.

    6/ In terms of favourite race/class combinations, we have the following:
    – AD: Goblins and Tigrans are both played three times. 1 race remains unplayed: Frostlings (it was played in other tournaments but it is indeed the least played race for AD in the total meta stats)
    – Dread: Dwarves keep dominating with 8 Dwarf Dread vs only 1 or 2 for other races. Surprisingly, Human Dread has been played only once (but notably it was by Jean de Metz on round 5!), probably because Human is better combined with other classes.
    – Necro: High Elf makes a noticed entry with first Hiliadan, then marcuspers using them to win their match, inspired by Jean de Metz in another game. But there is no clear dominance of one race over others.
    – Rogue: Human (5), Goblin (4) and Orcs (4) are the main picks. Surprisingly, the strong Draconian Rogue combo (with Backstabbing Flyers) is not represented at all, probably because Draconian is mainly chosen as Sorcerer. Dwarves and Elves are also not played as Rogues.
    – Sorcerer: Draconian Sorcerer continue to dominate (8 out of 13), with its top Apprentice and Shrine cost reduction in Racial Governance. Honourable mention for Frostlings (4 out of 13).
    – Theo: Tigrans and Halfling are not played. The two beefiest races, Dwarves (4) and Orcs (5) are well represented, together with Halflings (4), who love Chaplain.
    – Warlord: Orcs on top (4 out of 13) but no success at all (100% defeat, vs Rogue Frostling, AD Tigran, Sorcerer Elf, and Theocrat Human). Draconians are not played.

    7/ Surprisingly, the races with the highest win-ratio are now the most under-rated ones: Halflings (70%), High Elves (67%) and (less-under-rated) Draconians (62%)

    The AoW3 Wiki, including a video library!
    Improve AoW3 with the balance mod (presentation video)
    The Battlefield.com/AoW3 the community website for AoW3

    #271157

    El_Lobo1986
    Member

    Nice post, but as you said, in this tournament players were forced to pick several classes and races. I think we shouldn’t overrate it. I can confirm that an orc warlord is one of the weakest warlord picks. Warlord can’t grant a resistance aura and the orcs get a hell of no physical damage even if the units are well levelled. I tried it once and never again.

    Don’t get me wrong but I like the fact that Gab lost his first ranked match. It’s like seeing Roger Federer losing a match. That give me this feeling that “players of the century” are just normal guys at the end of the day! 😉

    #271180

    Kruel
    Member

    Thanks for the analysis, it’s interesting to see the combo tendencies within the tournament. However, as far as class and race strength, I think reading too much into such a small sample size amongst widely varying skill levels will generate a lot of statistical noise. For example, in Round 1 let’s say the top-ranked player is playing a true relative newcomer to the game. Short of real-life problems getting in the way, the top player will likely win about 100% of the time with ANY race and class combo. This means the new player will also get the loss, no matter the combo chosen. So data from this game will throw off any analysis of combo strength, especially so in this case with such a small number of games relative to the wide range of combinations.

    In fact, due to the length of PBEM games and size of the player base, the most meaningful data regarding class/race strength you could probably gather would be from asking opinions of top players, short of some hugely in-depth analysis of the classes and races.

    #271202

    marcuspers1
    Member

    Nice post and discussion, I will give my subjective feedback below.

    I agree with most of what Kruel said, especially in terms of “matching your class/race to the difficulty of your opponent” – no offense.

    If we take my route to round 8 of the tournament, I made a table of what class and race I wanted to play in the different rounds, and keeping in mind the relative strength of the opponent.

    So my list of classes in order of weak to best: DN, WL, Sorc, Rogue, AD, Theo, Necro

    In my first match I look to match the weakest class with the weakest race (of the 7 i need to use)

    Round 1 – Very weak opponent (no name needed, but if you are curious you can find the details!) I picked the weakest possible class (DN) with a weak race (Hafling) – they have some interesting treats as DN (apron wearing robot) but otherwise a calculated risk for a 1st round opponent. Opponent also play DN, so both had the same idea.

    Round 2 – Strong opponent
    I pick the 2nd weakest class (WL), but match it with an uber strong race (Tigran). I ended up losing this match, but I faced arguably the best class (necro) with a Frostling race. Opponent play his wildcard class, so a different strategy.

    Round 3 – Weak opponent
    I pick the 3rd weakest class (Sorc) combined with, for Sorc, a strong race, Frostling. Opponent play wildcard (necro)

    Round 4 – relatively strong opponent
    I pick the 4th best/worst class /Rogue) with a weaker race (orc). I rush bards/scoundrels to evolve and charm. Opponent play WL (Human) his favourite pick :party

    Round 5 – Strong opponent
    Now it’s time to use the top 3 classes! I again face the same opponent as in round 2, and I pick my go-to class/race: Human AD

    Round 6 – Very strong opponent (3rd ranked player at the time)
    I pick Draconian Theo – So the 2nd best class with a good race, had the option of Dwarf but opted for the range attacking evangelists (assumed opponent would play Necro). Unsure what opponent played, as he gave up at turn 7.

    Round 7 – Very strong opponent
    My last pick is Necro and I can choose between Dwarf, Elves and Goblins.
    I go for Elves because of the evolve strategy on the t1 irregulars (tournament is played with weak start so more often than not you get 3xt1 irr). Opponent plays Gob WL (he used all his strong picks in round 1-4 except Rogue)

    So, I think this shows the natural order and why some picks get a higher win ratio than others, simply due to the fact that you adapt your picks based on the strength of the opponent. It obviously gets harder at the later stages of the tournament and most players probably save the strong picks for those matches. So you will have 2 top players picking 2 strong classes and 1 have to lose.

Viewing 12 posts - 1 through 12 (of 12 total)

You must be logged in to reply to this topic.