Fall down 7 times, get up 8.
Do or do not, there is no try.
Don't make the same mistakes twice.
Insanity: Doing the same thing over and over again and expecting different results.
You can't always get what you want.
Better luck next time.
That's the stuff people tell you when you don't get the results you wanted. You are told, in various ways, that you have to pick yourself back up and get back on the horse. That you have to visualize success. That you have to learn from your defeat. I say, okay, let's do that, but in the crazy logic style of PKB.
I wanted everything. I got the silver and the gold. I could have played for another hour and a half and probably have gotten the gun, but after eight and a half hours of doing pretty much non-stop PVP (200 fights as it turns out), I couldn't really face the prospect of doing another forty to fifty.
There are a couple places where it went wrong.
@nickbrianlin We are aware of an issue where PVP Armory Bonus is not being calculated correctly and working on a fix to be released soon.— Justin Woods (@ProducerWoods)
Yep. That would be one of them. See, I was able to do something asymmetrical in the previous tournament, where I won a lot more AFK fights than I lost initiated fights. Basically, 82% of fights I started, were won, while 36% of fights against me were also won. Add those up and you get 118%. Now, due to asymmetry in the number of fights, it doesn't work to winning 59% of all fights. But the fact of the matter is that I won the top prize in the second tournament on the strength of my AFK results. I lost in this tournament on the weakness of my AFK results, which were the worst ever, despite running higher bonuses and what everyone said should be a strong team. I figured, based on what I was seeing, that something was wrong with bonus calculation, so I went looking and found the above tweet.
The other place it went wrong was in pattern recognition. That's on me. Since there was a perfect symmetry between my offensive results and my defensive results, the impetus should have been to put on an ungodly, unmatchable amount of volume. My belief is that you want the following numbers to total greater than 100%*:
- Initiated fight win percentage
- AFK win percentage
By getting those >100% collectively, you have created an imbalance that will result in a higher score. By getting it above 110% your score should be dramatically higher. I saw this happening in the later days of the tournament and did not do enough volume when I was winning a lot of fights, based on the amount of AFK volume I was seeing.
*There needs to be some parity in volume, obviously. 1 win in 1 intiated fight plus a 10-90 afk will not get you anywhere useful, obviously, but 369-79 in intiated and 202-422 in AFK would have won you Psylocke, as it did me.
As always, the goal is the top prize, while I don't like the hero (X-23 say my birdies... Clones are Stupid according to me and Tony Stark), I kind of have some unfinished business, right? The funny thing about the tournament, X-23, I have zero interest in any league above Diamond. But, for you, loyal readers, I'm gonna go for all rewards.
Somewhere between level 260 and 270, you start getting matched up against people at L300. Now, last tournament, I had a few fights against people 20 levels above me and, while the hero bonus is less important now and the armory is more important, I did terribly against the few people of much higher level and I'm not eager to be easy prey for people with 30 levels on me. As a result, it's time to strap some levels on, like 41 (to get from 259 where I finished the last tournament to 300, where I won't have the "higher level enemies" problem).
It took me 24 days to strap those levels on. I put up over 75 additional million on mission 10.6, which I will be happy do never do again, but I made it on March 1st. I thought it would take a week longer, but man, what you can do when you put your nose to the grindstone.
At the end of the last tournament, I had two heroes level to get to L12. Deadpoo who was needed in the Havok operation and Luke Cage, who was needed for running 10.6, the best XP mission in the game.
Of course, it's a moving target, so, as soon as I get Poo and Cage up to 12, I win Magneto and they release Wasp. And just to make the challenge of leveling Cage, Wasp and Maggy to 12 that much more challenging, Captain Britain was released in fairly short order. As of this writing (March 1st 8:32 AM), Poo and Cage are 12, Wasp and Cap B are 11 and Magneto will be 11 before I go to bed.
8 marginal wins. That's 100 points of ELO. Assuming we're being matched reasonably well, 64 wins above .500 should net me all prizes. In theory, you could win the top prizes on 100 attacks, as long as you didn't lose too much AFK or too many of your offensive fights. But, at this level, that won't happen and with the new daily roulette and the leveling bonus for the first five fights, I simply have to get in early and put some volume on. The goal is going to bet based around marginal wins. I want to be 8 more wins over .500 at the end of every day. That should give a buffer in the AFK that will enable me to start the final day somewhere above the silver league.
The Last Day
Do not think that I can expect lower volume on the final day. There are people with HUGE stores of challenge points, so there's no reason to expect anything different on this final day or the next final day. Maybe even not the final day after that.
I futzed around with various teams in the early goings until I settled on an offensive team of:
This is not a good defensive setup, as Widow is massively inconsistent. On offense, she's very strong, with buff removal, stun, counters, stealth and target spotting. Very strong. Emma, fairly obvious. Buff cleaner, damage reducer, stuns, psychics, and area attack. And the agent kit. Three counter attack weapons. Fury is great protection against Strange and Dark Sigil as it gets stronger, and the Signpost gets kicked up with Fury's Deathfrost. Wave Nullifier offers some additional buff removal and nerfs tank heroes, Quicksilver, Dr. Strange, Thor, Hulk, Magik, Gambit, and Warbringer Axe. It also does nasty things to Captain Britain (no Roars, but still gets extra turns), Nanite Nailgun (no shields), Psylock (no psychic shadow or shield), Emma (prevents application of Unlock Potential), and quite a few other buff dependant heroes and tools. Knight is what makes my agent a tank, has some synergy with Fury (except it always counters first... always Knight, Wavegun, Fury... would love it if it were Wavegun, Fury, Knight). This is solid and the type of team that I like to run... lot of options for what to do.
You will notice the lack of a QJ. I find QJ very annoying when other people play it now that I don't. So, don't carry yours in the first slot, else you will not get a chance to use it. I am prefering carrying the healing of the Signpost (and accepting the 10% damage reduction of poison for the large healing and accuracy improvement). I am not passing on QJ because it's useless or slow. I am passing because I like additional aggression. I think the QJ has some serious potential for "programming" the AI. Consider, something like Psylock, Captain Britain, with a ghost tanking agent. On the first turn, the agent pops a quick action item, and ghosts out. On the second, you have a 33% shot of the agent doing the correct thing (QJ) and a 67% shot of the wrong thing (an attack that would remove the Astral effect). It's probably more like a 60+% chance that he will use the QJ as it's slotted fairly high priority., then it's extra turns for Cap B (assuming he hasn't taken an extra) and Psy, with chances of coordination. Not bad. I might want to try this, actually. I do think this might be a nerf too far, but I see tremend
Has been an ongoing work in tweaking. Many possible strategies, but few that actually work. And I see my AFK number as not being that horrible, despite it's <30%. As I look at the other people around and check their profiles, I don't see too many with high AFK win percentages. So, maybe, I should just stick with something that produces the 30% and maintain with volume.
If I come across anything that works really well, I'll let you know.
I have a ton of data, courtesy of forum member Caslandr and his Random Fight Order Project. If you want to log like I do, there's a contact link on the RFOP page where you can sign up. He's looking especially for some <100 and <50 players to help out. It doesn't take that much extra time (maybe 10 seconds a fight, on the long end) to log with his tool and it's useful to you if you want to be a very competitive PVP player on a beer money or less budget.
The volume table was not tracking data at the start of the tournament, so you see the data on Feb 18 as if that were the start of the tournament. That is obviously not the case.
|Day/Time||Rating||Ses ATK||ATK Win||ATK Lose||Tot ATK||ATK W%||Ses DEF||DEF Win||DEF Lose||Tot DEF||DEF W%||Tot A / Tot D||Length|
If you're going to do well at AFK, you are going to have to know what they are playing. That's the metagame. The sum total of what everyone in the game environment is playing. Some quick notes about the metagame.
The meta is different at different levels. You will notice that I don't see much Human Torch, Dr. Strange or Invisible Woman. I know one player is complaining that those are all that anyone plays. I am not going to claim his data is false (it probably is exaggerated, but not outright false), but it proves that the metagame is different at different levels.
The metagame is also different at different tiers within level groupings. This should be fairly obvious, as better strategies will go up in ranking and lesser strategies will fall down. Volume matters in my meta, but it's harder to rise with a sub-optimal strategy. If you're not winning at least 80% of your offense, you are going to need to rethink your strategy somewhat, as that's not gonna get'r done.
Lastly, the metagame is never static. As trends set in, strategies evolve. People more experimental than I get to work on solving the problems of a particular meta with heroes (rather than tactics, which is my first solution). But the meta will move when new stuff comes in, or when dominant pairings settle in (q.v. me running the Fury... Lot of Dark Sigil out there, and why remove the debuff if I can make something very aggressive out of it?). So, things change. Faster at lower levels where there are more players. Be alert.
|Hero||Wins||Losses||Win%||Total||Share of Fights|
Well, I made it to Vibranium. I'll take it. Much better than last tournament. A lot like the last day of the second tournament. At any rate, I'll let the picture speak for itself:
What it hard to get from the picture is the final day volume and the shape of that. I should make a graph just for the last few hours. I went 0-45 in the last 3+ hours of the tournament on AFK. That was against 37-9 on offense. To put it simply, I knew that something like the 0-45 could happen, and I just didn't build it high enough to withstand a finishing kick of 37-54. I could have run more fights earlier in the day to build a larger buffer. And should have. But, since my only real goal with this tournament was to get to L300 and finish Diamond or better, I hit all my goals.
Going forward, I think Vibranium is my goal. I know, that's not that ambitious, but since the heroes come out a month later (getting my Punisher in 3.5 hours) and the uniforms come out never, I think I know the goal going forward.