Home Scoring and other IT questions

Extracting results from BBO

I have been extracting and uploading results from BBO as a playing Director without a problem. One goes into History, gets to the relevant game and does it. But the other day I didn't play and acted as a non-playing Director. At the end of the session when I tried to extract the result using BBO Extractor immediately after the game was over it came up with 'Unable to link with Travellers, giving up'. I tried again some days later as a non-playing Director and got the same answer.

Is there a way for a non-playing Director to extract the results and convert them to XML?

Comments

  • You should be able to extract results regardless of whether you're a non-playing director, a playing director, a player or just happen to be interested in a particular set of results. If the scores are provisional when you attempt to extract them it may come up with an error, especially if they convert to final during the process.

    There are other explanations though - if the results have been uploaded to BBO's hand records server, but the server is very busy then BBO Extractor may run out of attempts to connect to it at which point it will fail as with any program. Equally, BBO has problems during the evening on occasion with uploading lots of tournaments finishing at around the same time; Our club has had to wait until 6 or 7 in the morning to start processing results from the previous day for the last two weeks.

  • I had no problem running BBO extractor on a match that I had no involvement in as player or TD. I say no problem, there was some inconsistency in the output depending on how I generated the link.

  • I actually had this same error come up today but I realised I was using the wrong url. If your url is of the form http://webutil.bridgebase.com/v2/tview.php?t=, and the travellers have been uploaded to BBO, then it should work.

  • Well it seems that the BBO hand record upgrade has been very successful and results are now ready for extraction within seconds of each tournament ending. At evenings it still takes a bit longer than at other times of day but this is a massive improvement, so thanks first to the BBO tech team (who I suspect will never read this!).

    I've been using the Neuberg scoring feature in BBO Extractor so I haven't had a problem, but I've noticed a few oddities with the matchpoint column and the percentage column not matching up when basic scoring is used, for example...
    https://www.bridgewebs.com/cgi-bin/bwon/bw.cgi?pid=display_rank&event=20200623_1&club=cambridge (Tie for 2nd EW)
    https://www.bridgewebs.com/cgi-bin/bwon/bw.cgi?pid=display_rank&event=20200620_1&club=surrey (10th and 11th NS swapped, 10th and 11th EW swapped, 5th and 6th EW apparently not tied)

    In each event, there were some averaged boards which may have been the cause?

  • It's a consequence of the way that BBO calculates the percentages for the ranking list. It does so based on the percentages for each board after they have been rounded to two decimal places, rather than rounding only after the calculation has been performed. Thus, it's possible for two pairs to get the same percentage but with different matchpoints !

    If you don't select the Neuberg option, BBO Extractor will match the board scoring used by BBO, and will not modify the percentages assigned by BBO in the ranking list. That is, unless you use the new facility to edit a score in BBOtoXML v1.2.5 - if you do so the ranking list will be recalculated whether Neuberg is on or off.

    The reason it works this way is to make sure the published results match the results published on BBO in every respect if you select the "Scoring matches BBO" option. If you select "Enhanced Scoring (Neuberg)" then you will not see these anomalies between matchpoints and percentages on the ranking list, though of course the ranking order may be different to what is shown on BBO.

  • edited July 2020

    Just a minor thing about Neuberg scoring and 60/40 scores. It looks like WB 4.1.1.1 (copied below) isn't currently used:

    "However, in a match-pointed pairs event, if a pair’s average score (percentage) is greater than 60% then they get their average score for AVE+. Similarly, if a pair’s average score is less than 40% then they get their average score for AVE−. A similar approach is used at IMPs, e.g. a pair that has an average of +4 IMPs gets +4 IMPs for AVE+, and a pair that has an average −4 IMPs gets −4 IMPs for AVE−."

    Obviously in BBO events a 60/40 score is relatively rare, but may occur if, for example, a player has connection issues but their partner would instead prefer wait for them to come back online (and is in contact with their partner by phone or similar) as opposed to having a robot sub in the meantime.

  • @495670 said:
    Just a minor thing about Neuberg scoring and 60/40 scores. It looks like WB 4.1.1.1 (copied below) isn't currently used:
    "However, in a match-pointed pairs event, if a pair’s average score (percentage) is greater than 60% then they get their average score for AVE+. Similarly, if a pair’s average score is less than 40% then they get their average score for AVE−. A similar approach is used at IMPs, e.g. a pair that has an average of +4 IMPs gets +4 IMPs for AVE+, and a pair that has an average −4 IMPs gets −4 IMPs for AVE−."

    Obviously in BBO events a 60/40 score is relatively rare, but may occur if, for example, a player has connection issues but their partner would instead prefer wait for them to come back online (and is in contact with their partner by phone or similar) as opposed to having a robot sub in the meantime.

    What makes you think it isn't used?

  • @495670 said:
    Just a minor thing about Neuberg scoring and 60/40 scores. It looks like WB 4.1.1.1 (copied below) isn't currently used:
    "However, in a match-pointed pairs event, if a pair’s average score (percentage) is greater than 60% then they get their average score for AVE+. Similarly, if a pair’s average score is less than 40% then they get their average score for AVE−. A similar approach is used at IMPs, e.g. a pair that has an average of +4 IMPs gets +4 IMPs for AVE+, and a pair that has an average −4 IMPs gets −4 IMPs for AVE−."

    Yes, that's true, BBOtoXML currently assigns a 60/40 score as 60%/40%, even if the pair's percentage is above 60% or below 40%. We'll change it so that it's done according to WB 4.1.1.1

  • edited July 2020

    @495670 said:
    It looks like WB 4.1.1.1 (...) isn't currently used:

    WB 1.1.1.1 is only a restatement of Law 12C2 (c)

    The foregoing is modified for a non-offending contestant that obtains a session score exceeding 60% of the available matchpoints or for an offending contestant that obtains a session score that is less than 40% of the available matchpoints (or the equivalent in IMPs). Such contestants are awarded the percentage obtained (or the equivalent in IMPs) on the other boards of that session.

  • I ran a session yesterday and extracted it in the normal way, and just happened to notice it in the travellers when I uploaded the results to Bridgewebs. I thought it was probably in the Laws somewhere but since it was a scoring-related query I tend to look at the White Book first - thanks Robin for pointing that out.

  • In working on this issue today I was comparing my results with those produced by BBO Scorer as a sanity check. I came across a small bug in BBO Scorer. This can be seen in the traveller for board 6 of the EBU Game 3 event of 6th July 2020. NS pair 69 played against EW pair 110. the contract was 1NTxx by E, making 6 tricks. The declarer is shown as "X" rather than "E" and the score is shown as 200 (should be 400, as EW are vulnerable). I assume BBO Scorer is interpreting the second "x" in "1NTxx" as the declarer.

  • @jgoacher said:
    Yes, that's true, BBOtoXML currently assigns a 60/40 score as 60%/40%, even if the pair's percentage is above 60% or below 40%. We'll change it so that it's done according to WB 4.1.1.1

    Hi and sorry for bringing this up again. I ran an event today which had a 60/40 score... the Av+ pair scored 50% overall while the Av- pair scored 61% overall. BBO Extractor changed the result to 60/39, so it looks like the 60/40 correction has a small bug? This made a difference of 0.08 matchpoints out of several hundred so it's not very important but I thought you should be made aware.

  • Could you post the BBO tournament link and I'll check it out.
  • Thanks for spotting this. I've fixed it now in BBOtoXML. You will need to refresh the BBOtoXML web page in your browser, or perhaps clear cache, in order to get the updated version.

  • Thank you John for your willingness to respond so quickly to feedback like this.

  • @gordonrainsford said:
    Thank you John for your willingness to respond so quickly to feedback like this.

    No problem, it was a simple error in one line of code and easy to correct.

  • The club haven't yet uploaded the new results but I'd like to second Gordon in thanking you for the very quick update, both for this and the initial enquiries.

  • Following system updates / reloads last night ... Chrome /Chromium is blocking the download of .pbn files. There doesn't seem to be a way to accept the risk, just "this was blocked; click to discard the file". Happens on both Windows10 laptop and RaspberryPi 4B. Anyone having the same issues / know how to fix this?
    Peter

  • @pg10003 said:
    Following system updates / reloads last night ... Chrome /Chromium is blocking the download of .pbn files. There doesn't seem to be a way to accept the risk, just "this was blocked; click to discard the file". Happens on both Windows10 laptop and RaspberryPi 4B. Anyone having the same issues / know how to fix this?
    Peter

    There is a way to keep the file. I've just sent out this email to people on our contacts list:

    "For reasons best known to themselves, google have started blocking access to PBN files downloaded using the chrome browser. This started this afternoon, and of course includes the PBN files created by BBO Extractor. It appears to occur for PBN files downloaded from any source, so BBO Extractor hasn't been singled out for special treatment ! It also occurs for LIN files downloaded from BBO.

    You will see a message at the bottom left corner of the browser windows stating " may be dangerous, so Chrome has blocked it" and a "Discard" button next to it. To get access to the file, click on the "Show All" button at the bottom right of the browser window. This produces a list of files in the Downloads folder. Then click on the "Keep Dangerous File" button for the particular file entry in the list, and finally click on "Keep Anyway" when you are asked to confirm. Phew !

    There are a couple of other alternatives:

    1. Turn off safe browsing protection in the Chrome settings. I wouldn't recommend this, as it exposes the PC to genuine security risks.
    2. Use BBO Extractor with a different browser that supports Chrome extensions. On the PC this includes the latest version of Microsoft Edge, also Opera, and Yandex."
  • I had this problem in my afternoon game. The .pbn file was in my downloads folder but with completely the wrong name. However when I ran these evenings game through the extractor everything was fine and others have reported the same.

Sign In or Register to comment.