Records
Event Records
2020 Final Scores (@Home)
2019 Final Scores (BSDE)
2019 Final Scores (BSDC)
2019 Final Scores (BSLV)
2018 Final Scores (BSDE)
2018 Final Scores (BSDC)
| Team Name | HingleMcCringleberry | MeeseeksAndDestroy | PopPopRet | RaidersOfTheLostArp |
|---|---|---|---|---|
| Team Score | 652,282 | 778,136 | 615,063 | 874,927 |
2018 Final Scores (BSLV)
| Team Name | KnightsWhoSayNi | :(){ :|:&};:: | PaisleyScratcHMonkeys | ArcanumExNihilo |
|---|---|---|---|---|
| Team Score | 118,698 | 100,407 | 94,779 | 258,884 |
2018 Final Scores (HackWest)
2017 Final Scores (BSDE)
2017 Final Scores (BSLV)
2016 Final Scores (BSLV)
Final Scores
| Team | Infinite Improbability | The SYNdicate | Team Machine | JMP-ESP |
|---|---|---|---|---|
| Day 1 total | 17,860 | 20,788 | 20,943 | 19,886 |
| Day 2 total | 3,114 | 20,148 | -299,571 | -24,335 |
| Grand Total | 20,974 | 40,936 | -278,628 | -4,449 |
-
The 2016 Scoring Bug
This year, we had the misfortune to deal with a critical bug in Scorebot, the program that we use to run the ProsVJoes CTF. This bug seriously impacted the scores displayed on the game board, which was inaccurately representing the state of the game for both days.
Right up front, I would like to personally apologize to all of our players for this flaw, and for the ill feelings this may have caused. I have never seen a Pros V Joes CTF where so many people put in so much effort for so long before the actual game. It is only natural that these fine people have an emotional investment in the game, and so may have been extremely disappointed in the final results once they were published on Twitter.
Unfortunately, the bug is what is, and I can only do the best I can to make it right and fair after the fact, now that this year's game has concluded.
As a small mercy, the bug was isolated in scope to only the ticket scoring portion of the game. The rules state that players must close tickets as a part of the game. To score them on this, the scoring software monitors the open and closed tickets. The design had been that the open tickets a Blue Team had would cost them 50 points per service scoring round (about 3 minutes, on average).
Here's where the bug comes in.
Scorebot was docking each team 50 points per service round for every closed ticket.
The bug was discovered on day two by one of our volunteers this year, Gambite, so a big shout out to him.
Unfortunately, no way exists to handle this bug during run time. Once Scorebot starts, that's it, she's off and running for the duration. So our only choice was to fix things after game play stopped.
For the second day, we had the entire ticket database - every flag, when each was opened, and when each was closed. So, we could go back and calculate what the final ticket score was for day two.
For the first day, we had no such advantage. Historically, we wipe the ticket database at the start of day two, so that the prior day's tickets are no longer counted, since each day's game play is independent. While scorebot generates a great deal of log data to track almost every event in the game, the ticket code does not currently log open / closed tickets for every scoring round. That's bug number two. This was a simple, yet damaging oversight in the code base that will be corrected before the next game.
So, in the end, we were able to recalculate the correct scoring for day two, but we did not have the data needed to do the same for day one.
So what did the numbers look like for day two? Here's what they were, and what they changed to:
Team Name Infinite SYNdicate Team Machine JMP-ESP Old Tickets -198,000 -156,350 -63,650 -297,350 New Tickets -4,250 -14,900 -314,750 -14,000 As can be seen, this is a drastic change. Again, it is a consequence of the bug counting closed tickets as open when calculating the amount of points to take away each service round.
As a result, here's what day two looks like after taking these new scores into account:
Team Name Infinite SYNdicate Team Machine JMP-ESP services 20,464 20,648 19,779 20,165 flags 600 15,400 7,600 -3,500 tickets -198,000 -156,350 -63,650 -297,350 beacons -13,700 -1,000 -12,200 -27,000 Old Total -190,636 -121,302 -48,471 -307,685 New Tickets -4250 -14,900 -314,750 -14,000 New Ticket Total 3114 20,148 -299,571 -24,335 For day one, we did not have the data needed to make that sort of a recalculation. As a result, our only options were to count the data as is, or discard that part of the score entirely.
We chose to discard the tickets, which seemed the most fair to the teams that had been diligent in closing their tickets. Here is what day one looks like, with that choice:
Team Name Infinite SYNdicate Team Machine JMP-ESP services 21,360 20,588 19,943 20,086 flags 400 1,400 2,100 2,600 tickets -94,800 -100,950 -64,950 -104,750 beacons -3,900 -1,200 -1,100 -2,800 total -76,940 -80,162 -44,007 -84,864 total w/o tickets 17,860 20,788 20,943 19,886 So, adding up the adjusted first and second day scores as described above, we come up with the results that were published on Twitter:
Team Name Infinite SYNdicate Team Machine JMP-ESP Day 1 total 17,860 20,788 20,943 19,886 Day 2 total 3,114 20,148 -299,571 -24,335 Grand Total 20,974 40,936 -278,628 -4,449 I have personally had multiple conversations with various Blue Pros and Blue Joes from multiple teams on this topic. Thus far, everyone I have explained the circumstances to was understanding about the difficult choices that had to be made. Indeed, three members from Team Machine, undeniably the group that was most impacted by this issue, have agreed to come on staff in spite of this snafu.
Regrettably, my team and I can do nothing about this issue beyond what the solution just described. What we can do, however, is work hard to ensure our future games in the remainder of this year and into next do not suffer from this, nor from any other such egregious flaw. To that end, we're re-doubling our efforts to clean up the existing 2.x code train, while also restarting effort on the new 3.0 architecture. We will be doing more code reviews and testing, working hard to do our best to prevent a recurrence. Of course, mistakes will happen, but we intend to learn and improve from this one.
To that end, I am pleased to announce that Gambite, the volunteer who found the bug and helped me calculate the correct scores, has signed on to our permanent staff as part of the Pros V Joes dev team.
2015 Final Scores (BSLV)
The 2015 Pros V Joes CTF was our largest yet, with 44 players and 18 volunteers contributing to make the biggest CTF BSLV has ever seen! Everyone participating did remarkably well. Thanks all for playing, and I hope to see you next year at BSLV 2016!
| Team | Score |
|---|---|
| Endtroducing | 8,862 |
| Labrynth Guardians | 33,785 |
| Castle Keep | -20,781 |
| Salty Goats | 1,831 |