Daily Archives: May 3, 2023

Looking at Destruction as a Metric for EVE Online

Last week, in my look at destruction in EVE Online for March 2023, I kind of hit on the idea that average destruction per day over a month might be a reasonably metric to keep an eye on.  It is an easy enough calculation, something even my meager Power BI skills can gin up to be calculated automatically.

My thought, when I wrote that, was that it would be something I could test over time.  But then I realized I most of the past MERs sitting in folders on my computer, so rather than looking forward I could check into the past and see how they played out against what I though was going on at the time.

Easy enough.  There would be some tedium in setting it up, but once done the data would be there.

Then I started thinking about how to make the assessment.  It is fine and dandy for me to SAY it looks good.  Oh yeah, COVID led to a big spike in users in March/April of 2020, so numbers should go up.  Likewise, when summer 2021 hit everybody was done staying home, bored with the stalemate of World War Bee, and keen to go outside, so we had the great June/July dip.

But it would be nice to have a bit more than that, so I went over to the charts that Jester keeps up to date to grab the one that shows the average logged in players.  That would work as something to measure against.

Average Players Logged in Over the Last Four Years

I went and put the data together in a quick table.  The calculations are fine, though there was a flaw in the MER up until late 2021 or early 2022 where CCP was only counting the first 30 days of the month, so anything that happened on the 31st was left out.  That mostly doesn’t matter, save for Dec 31, 2020, which was the day that the first battle at M2-XFE happened and 257 titans were destroyed.  So went back and chopped a day off of all months with 31 days in 2020 and 2021.

This is the data I rolled up from that.

2020 2021 2022 2023
Jan             12,609.17              16,223.17           14,035.39           13,729.97
Feb             12,136.41              15,307.64           13,331.61           14,307.75
Mar             21,308.93              16,222.73           11,773.29           14,878.68
Apr             20,564.73              13,509.93           12,343.07
May             16,154.60              13,661.17           11,364.16                          –
Jun             15,420.10              11,646.27           11,857.00                          –
Jul             15,930.83              11,942.37           11,719.68                          –
Aug             15,791.47              13,638.03           11,617.29                          –
Sep             14,709.83              12,899.03           11,195.90                          –
Oct             15,880.03              13,010.40           11,172.87                          –
Nov             14,945.63              13,099.40           14,148.90                          –
Dec             14,090.70              12,709.60           13,412.77                          –

Put into a graph to try and line up to what Jester has recorded got me this.

Average Losses per Day

That lines up fairly well.  Jester’s chart is smoother because he does a daily calculation, and it is also a bit out of sync because it is a 30 day rolling average, so changes show up a little later.  Not perfect, but it matches as closely as the difference in data probably allows.

So it seems like it could be pretty good for a month to month thing.  Now, if I was ambitious I would go in and Power BI do me up a chart that did daily average, but that is more work than I want right now so I will leave it be.

However, having done that, I was wondering if any other metric might follow the same lines as closely.  So I swapped out for average daily losses by ISK value rather than just the raw total, a data set that looks like this.

2020 2021 2022 2023
Jan 1220.22 billion 1797.53 billion 1085.69 billion 1325.54 billion
Feb 1252.76 billion 1279.04 billion 1039.56 billion 1248.78 billion
Mar 1440.87 billion 1364.97 billion 916.49 billion 1271.91 billion
Apr 1471.13 billion 1111.62 billion 986.45 billion
May 1432.14 billion 1054.00 billion 988.82 billion
Jun 1459.95 billion 898.98 billion 922.96 billion
Jul 1124.64 billion 885.27 billion 927.13 billion
Aug 1164.18 billion 1141.21 billion 857.03 billion
Sep 1073.76 billion 1022.62 billion 940.40 billion
Oct 1379.73 billion 1033.09 billion 967.36 billion
Nov 1245.81 billion 1013.29 billion 1065.24 billion
Dec 1091.41 billion 1061.98 billion 1190.89 billion

That gave me a chart that looked like this.

Average Daily Losses by ISK value per month

That is a little more flat.  You can still sort of see the patterns that Jester’s chart displays, but they aren’t as obvious and some points, like January 2021 where the second battle of M2-XFE happened and was actually recorded, stand out much more dramatically than perhaps they ought.

So that chart seems a little less useful as an overall metric, if only because it is, as one might expect, influenced by expensive battles.

Finally, I decided to see if I could combine the two sets to get a third chart, which gave me the average cost of losses over the course of the month… basically the average ISK loss divided by the average losses per day.  Probably not the best data set, but it was easy enough to throw together with the work I had already done.

2020 2021 2022 2023
Jan 96.77 million 110.80 million 77.35 million 96.54 million
Feb 103.22 million 83.56 million 77.98 million 87.28 million
Mar 67.62 million 84.14 million 77.84 million 85.49 million
Apr 71.54 million 82.28 million 79.92 million
May 88.65 million 77.15 million 87.01 million
Jun 94.68 million 77.19 million 77.84 million
Jul 70.60 million 74.13 million 79.11 million
Aug 73.72 million 83.68 million 73.77 million
Sep 73.00 million 79.28 million 83.99 million
Oct 86.88 million 79.40 million 86.58 million
Nov 83.36 million 77.35 million 75.29 million
Dec 77.46 million 83.56 million 88.79 million

Average cost per loss per month

This chart looks nothing like Jester’s, which is not unexpected.  The ups and downs look perhaps more dramatic than they ought to because I narrowed the high and low range to be displayed on the chart in order to see some variation more clearly because, as it turns out, the average cost of ships lost in a month doesn’t change a lot from month to month unless something big is going on.

Most months the average loss is between 70 and 90 million ISK in value.  You need a big fight… once again you see what the second M2-XFE battle did to the chart… to give it much of a bump.  And even that only puts it up a little past 110 million ISK per loss.

When the losses are in the range of 400K a month total, things get averaged out pretty well, especially when there are so many low value capsule and corvette losses.

Finally, just because I was doing these numbers, I put together the annual totals as well as the average losses per month for each of the years on the chart.


  • Total Losses – 5,674,137
  • Total Value Lost – 459.45 trillion ISK
  • Avg Losses per month 472,845
  • Avg Monthly Loss value 38.29 trillion ISK


  • Total Losses – 4,899,115
  • Total Value Lost – 408.49 trillion
  • Avg Losses per month – 408,260
  • Avg Monthly Loss value – 34.04 trillion


  • Total Losses – 4,497,590
  • Total Value Lost – 361.49 trillion ISK
  • Avg Losses per month – 374,799
  • Avg Montly Loss value – 30.12 trillion ISK

2023 so far

  • Total Losses – 1,287,485
  • Total Value Lost – 115.49 trillion ISK
  • Avg Losses per month – 429,162
  • Avg Monthly Loss value – 38.50 trillion ISK

And that is what I did with my Friday night after my wife went to bed.