Hearthstone

[Kuro]

I managed to go only 2-3 with the following Deck:

Wow, that deck looks just like the one that usually beats me and ends my arena runs so who knows. Maybe you just got unlucky? The matchmaker seems to love trying to give 2 losses early by pairing against extremely good decks and then easing up a little until 5 wins or so. So maybe you just got a bad draw in the crucial third loss.

What was beating you? Getting swarmed in the early game by stuff too big to consecrate, or were you losing late game draw-outs?

My matches were:

Loss vs. Paladin, he just outplayed me and more importantly had Aldor Peacekeepers.
Loss vs. Hunter, didn't draw any of my power cards and lost.
Win vs. Mage
Win vs. Mage with a really weird Deck that was like the super stall Constructed mages but since it was Arena he did like nothing.
And then a Loss vs. a Mage where I got totally crushed because Consecration only does 2 damage so I was totally owned in the mid game.

[Gustaran]

With regard to Gustaran's point, I think the hammers of wrath are some pseudo-lategame, and I think you may have had a pretty good deck there that just got unlucky. I think my times I've gotten less than 3 wins have been more about unlucky games (opponent curving out, you flopping) than anything else.

Ah yes, Hammer of Wrath indeed does add to your cycling through the deck. I fail to see, however, how a Hammer of Wrath adds a lot to your lategame? If your opponent plays an Ironbark Protector and you draw a Hammer, it's not exactly enhancing your game by a lot.
I agree on the luck part, though. Keep in mind that for every win you get, somebody else needs to lose. So if you are up against decent opposition you might very well only go something like 2-3.

[SevenSpirits]

Well sure, it's not specifically an answer to Ironbark Protector on its own. But it sure helps you spend all your mana.

Suppose there were a card that does 5 damage for 4 mana. (That's completely in line with how other single-target damage spells are costed, except of course the mage ones are 1 tick better.) Then you could hammer of wrath the protector, draw the made-up card, and finish it off, for the same mana and card cost it took to play that protector. Now obviously that's both impossible and unrealistic, but the question isn't whether it does anything against protector in particular. It's about whether you get a comparably powerful play from it, and I think the answer is yes. Average cost in your deck is probably close to 4, so it's eating up 8 mana and providing 8 mana worth of strength. And while this may end up being a bit less strong than protector, this card is much more flexible: you can play it along with a variety of other things, not just as a single unit for 8 on the same turn. So I definitely see it as a good card to draw late-game.


The win-loss system they have for arena is good for strong players, but must be painful for bad players. Something that might not be immediately obvious is that the median record is (with good matchmaking) 2.5 wins and 3 losses. Literally half the runs people play never reach a 3rd win. (Suppose after four rounds, some number of people are 2-2. Some did worse and are out, and an equal number did better and are at 3+ wins. In their fifth game, half the players lose and are out. In total, 50% get 2 wins or worse, while the other 50% get 3 or more.)

But back on topic, I suspect Kuro is better than the median.

[Dantski]

If we're talking bad plays, today I had a situation where I needed to do 6 damage to an enemy minion, and had a four attack minion active on the board and a stormpike commando in my hand. Turns out the best play isn't to drop the commando first and then click my active minion thinking I had dropped a dark iron dwarf

Heh that gave me a chuckle.

Main thing I have to remind myself is that I can silence my own minions and remove freezes and debuffs.

[WilliamLP]

If we're talking bad plays, today I had a situation where I needed to do 6 damage to an enemy minion, and had a four attack minion active on the board and a stormpike commando in my hand. Turns out the best play isn't to drop the commando first and then click my active minion thinking I had dropped a dark iron dwarf

Heh that gave me a chuckle.

Main thing I have to remind myself is that I can silence my own minions and remove freezes and debuffs.

Especially Aldor Peacekeeper! That one isn't so obvious to me.

[Kuro]

Well sure, it's not specifically an answer to Ironbark Protector on its own. But it sure helps you spend all your mana.

Suppose there were a card that does 5 damage for 4 mana. (That's completely in line with how other single-target damage spells are costed, except of course the mage ones are 1 tick better.) Then you could hammer of wrath the protector, draw the made-up card, and finish it off, for the same mana and card cost it took to play that protector. Now obviously that's both impossible and unrealistic, but the question isn't whether it does anything against protector in particular. It's about whether you get a comparably powerful play from it, and I think the answer is yes. Average cost in your deck is probably close to 4, so it's eating up 8 mana and providing 8 mana worth of strength. And while this may end up being a bit less strong than protector, this card is much more flexible: you can play it along with a variety of other things, not just as a single unit for 8 on the same turn. So I definitely see it as a good card to draw late-game.


The win-loss system they have for arena is good for strong players, but must be painful for bad players. Something that might not be immediately obvious is that the median record is (with good matchmaking) 2.5 wins and 3 losses. Literally half the runs people play never reach a 3rd win. (Suppose after four rounds, some number of people are 2-2. Some did worse and are out, and an equal number did better and are at 3+ wins. In their fifth game, half the players lose and are out. In total, 50% get 2 wins or worse, while the other 50% get 3 or more.)

But back on topic, I suspect Kuro is better than the median.

Well, over 42 Arena runs that have been recorded and 299 matches, I average 4.12 wins per run and have gotten to 7+ wins 11 times (21% of the time), I think my "true talent" level is probably around 4.30 wins per run (Which is where it was for quite a while).

[Kuro]

Also, NH, something I thought of may be useful for your sheet, though I have no idea how to use it: It might be good to have it record the number of times you get to 7+ with each class as well as overall. This would help tell you how each average breaks down (A 6 average from going 12 wins and 0 wins alternately and 6 wins every time is much different to use an extreme example)

Also, how do I add more columns that are coded like your spreadsheet is in the spreadsheet? In case I ever get to the max of around 100 you seem to have now. Is there, like, a copy button?

[Cyneheard]

Also, NH, something I thought of may be useful for your sheet, though I have no idea how to use it: It might be good to have it record the number of times you get to 7+ with each class as well as overall. This would help tell you how each average breaks down (A 6 average from going 12 wins and 0 wins alternately and 6 wins every time is much different to use an extreme example)

Also, how do I add more columns that are coded like your spreadsheet is in the spreadsheet? In case I ever get to the max of around 100 you seem to have now. Is there, like, a copy button?

I think this was addressed to me.

Yes, I did set it to run up to 98 runs (Row 100).

Things you'd have to change to make it bigger:
- Under Formulas: Name Manager, for each of the named ranges make sure that the first number is 3, and the second number is 1000.
- Do the same for Home: Conditional Formatting, under Manage Rules. Make sure you select "This Worksheet" while you're looking at 'Runs'; there should be 14 different things you have to tweak.
- Select a row, say 5, in Runs, and then use Format Painter to have the formatting go from Row 100 to Row 1000
- In the Summary sheet, use Find-and-Replace to replace all of the "$100" with "$1000"; make sure that you check the option to search in formulas.

I didn't do the breakdown of runs-to-7+ by class, but I'd think that the "Ave. # of wins with" in D9:D18 are going to give you essentially the same information: I think "this class does very poorly in getting that 7th win, but does fine otherwise" falls under "random chance" and not "something's really going on here". I'm a little leery of cutting the data in too many different ways because the datasets are so small. I've only got 43 runs, 39 of which are well-documented, myself: barely enough to be confident about any conclusions I make about my play.

If we had thousands of arena runs, I'd try to do some statistical significance tests and the like to see if some of the trends we see are meaningful.

Also, have you ever used pivot tables? Always a good trick to have up your sleeve if you want to look at the data in different ways.

[Kuro]

Yes, sorry, I dunno why I said NH.

No, I've never used pivot tables. To be honest, I've never even opened my Excel before I got your Spreadsheet, so I'm very new to all this. Thanks for telling me how to expand it, though. :D

[Sir Bruce]

You guys should check out HearthTracker: http://blog.hearthtracking.com/. It runs in the background and automatically tracks a lot of stats.