USA Fencing SEMI Statement Regarding Saber Machine Timing

Discussion in 'Fencing Discussion' started by mtwieg2, Mar 22, 2018.

  1. Alex_Paul

    Alex_Paul Podium

    Joined:
    Jul 26, 2004
    Messages:
    1,079
    Likes Received:
    128
    What is interesting and surprising is that good fencers can tell a single millisecond of difference. We have done a lot of testing here with wireless recently and this has brought up some interesting information.

    In foil for example acceptable contact times for a valid hit are 13 to 15 ms fencers can tell the difference between 13, 14 and 15 ms contact (flick) times and prefer those that are easier to flick with.

    Further sabreurs (saberists to most of you ;-)) prefer ANYTHING that brings up more lights. If you reduce double time, remove the blocking or do anything that makes doubles more regular then they prefer it.

    Double time is 120ms +/- 10ms and the difference between 110ms and 130ms is pretty huge.

    Whipover in sabre has a huge acceptable window as it is 15ms plus or minus 5ms meaning after blade contact blocking of the next hit can be between 10ms and 20ms which is a massive window.

    If you set sabre up with the double time at its shortest at 110 ms and the blocking period at 20ms making the box its strictest allowable then fencers hate it and practically refuse to use it.

    Further to this having tested whipover extensively on a range of scoring machines there is very little consistency in terms of how it is implemented meaning that different machines that are supposedly to current rules are not correct. Most fencers (including Olympians!) actually don't seem to understand what whipover is supposed to do and I have had people demonstrate no light after a parry to me as a fault when it is in the rules.
     
    prototoast and oiuyt like this.
  2. oiuyt

    oiuyt Podium

    Joined:
    Apr 26, 2000
    Messages:
    10,268
    Likes Received:
    997
    You're cheating! You're running actual tests with real fencers to inform discussion of potential technical changes! That's not allowed!

    B
     
    InFerrumVeritas and ReadyFence like this.
  3. brtech

    brtech Podium

    Joined:
    Nov 4, 2003
    Messages:
    2,338
    Likes Received:
    175
    The sound test is NOT VALID. Whether the machine makes a sound when powered on in Saber DOES NOT tell you whether it's good or bad. There is some variation in the programming on specific series of machines that causes this variation but that doesn't tell you if the programming is updated or not. Some machines that pass the sound test have old timing. Some machines that fail the sound test have new timing. There is no simple way to tell if a machine is good or bad.

    The current suspicion is that there is something wrong with the programmer Absolute used to perform upgrades on these machines. There is something highly suspicious about what is reported about this. All programmers I've ever seen do a checksum of the content of the memory after programming to make sure the programming was completed correctly. It should have been obvious that there was a problem.

    I am told that US Fencing machines are not affected by this problem. All the machines we have are different series from the bad ones.
     
  4. Alex_Paul

    Alex_Paul Podium

    Joined:
    Jul 26, 2004
    Messages:
    1,079
    Likes Received:
    128
    :oops:
     
  5. 22hansen22

    22hansen22 Made the Cut

    Joined:
    Oct 8, 2009
    Messages:
    75
    Likes Received:
    0
    When this issue arised a while ago, I developed a testing machine that makes sure the scoring machine is within the tolerances specified by the FIE,
    it wasn't my first intention to commercialize this testing box but if there is an interest on it, send us an email and will work with you (more accurate than dropping sabers at the same time ;) )
    Cheers!
     

Share This Page