Quote from: eleuthria on June 18, 2011, 04:10:46 AM

Quote from: Veldy on June 18, 2011, 03:56:36 AM

Quote from: eleuthria on June 17, 2011, 07:03:24 PM

Quote from: smackdaddy on June 17, 2011, 07:00:53 PM

Quote from: eleuthria on June 17, 2011, 06:57:56 PM

Luck statistics were added to the block statistics page, for both a 24H window and since the difficulty change.

What do the percentages indicate? Variance from expected?

The percentages are related to the average shares per block in the past 24 hours/since difficulty change. Red numbers = bad luck, we're taking longer than average. Green = good luck, we're solving faster than average.

Just a little nit. The stats, at first glance, appear as if the blocks themselves are larger or smaller than difficulty would suggest [which happens]. One might read from the page that the blocks are larger than normal [when red/negative %], but it seems that you are indicating the number of shares submitted for the number of blocks earned won over a period of time? In that case, it still perplexes me, as the number of shares per solution should largely be the difficulty divided by the percentage of total network hashing power that BTC Guild wields. What am I missing?

To be fair, the same question can be asked of Deepbit too [I sort of always thought it indicated that larger blocks or smaller blocks than difficulty were been doled out for mining].

The more stats the better!

Understanding them exactly is an added plus [I have been working 14 hours straight today and my mind could be a bit addled].

The 24 hour number is calculated by taking the average shares it took for each block solve that was done in the last 24h. So if the block's solve time was within 24 hours, it's share count is included to determine the 24h luck figure. So it's not exactly 24h, since a block may have been solved 23h 59m ago, but it was started an hour before that. However, this number gives a representation that is fair in my opinion. The comparison is to the current difficulty to determine if we were ahead or behind the average. It is our luck relative to difficulty, not our luck relative to overall network block production in the last 24 hours.

I looked back and clearly I wasn't thinking. The pool hash rate (as a percentage of the network or not) has nothing to due with luck ... luck doesn't have a rate, so clearly I wasn't thinking when I asked that question (yesterday was a long day of debugging code and addled my brain apparently Sad

). It only makes sense that on average, if a round occurs entirely within the same difficulty period then on average, a block should be found in the number of shares represented by the difficulty number. I have no clue why my brain didn't process the obvious yesterday even being mentally fatigued.

Thank you for the stats; they are useful (in an academic sense) and illustrates a sort of one glance summary of the block statistics.