I’m not going to try to convert you, but I’ll point out that there have been a number of exchanges here to the effect of “The USCF says to do it that way, so that’s what everybody ought to do.” The USCF has no secret cache of knowledge; it’s made up of various individuals who don’t know any more about the subject than the people running tournaments. In a lot of cases, they know less. What I’m concerned with is a reaction like, “We don’t need to add [optional feature that would be good for the players]. The USCF didn’t include it, so it can’t be important.”
Here’s what I think Allan is trying to say (politely correct me if I’m wrong, not trying to goad anyone): the perceived need is a low-cost method of producing rulebook-approved pairings. He sees online technology and online availability as a way of making that happen.
I find the whole idea intriguing, and would rank it somewhere near the Atomic Bologna Slicer: serves a useful purpose, but does little more than currently available tools. The “gee-whiz!” factor would appeal to some, particularly those interested in web applications.
From a business analysis standpoint, the costs far outweigh the benefits of the USCF building the system. The political and human resource expenditures are significant when you consider the trade-offs among projects, not to mention monetary and physical resource cost (and then maintenance costs, lack of USCF budget, etc). Commercial apps (WinTD, etc) or low-cost resources (index cards) cover the spectrum of needs, and the TD “market” isn’t large enough to require another option.
If someone ever builds it I’d be happy to try it out.
I don’t argue with your executive summary at this time. But there are still good reasons to consider this idea now. (1) Can’t legitimately reject an idea without considering it first. (2) Would be nice for the USCF to be proactive for a change. The costs vs benefits will definitely shift over time, just a matter of how far how fast. (3) Not every project has to “break-even” for the USCF to undertake it, provided it is consistent with the mission. Although with the current financial state break-even is a requirement, this may relax in the future.
As for the “gee whiz” factor, I don’t care about that at all. I’m very boring and conservative in how I spend my nickels. But you can’t skrimp on the IT, because lack of productivity on the individual level really mounts up on the organizational level. The thing that hurts most about IT spending is it costs big bucks up front, but you can only measure the savings afterwards. Sometimes not even then.
Finally, just wondering: If I wanted to be an assistant TD at one of John Hillery’s events, would I be able to use SwisSys 9.6 (DOS)? Because that’s what I have on my computer.
-
As one who has proposed (small) IT enhancements in the past, I’ll say these forums will give you an unvarnished view of consideration. There’s a limit to the viewpoints: you’ll get opinions but no momentum. The forums are good to work out any bugs before an actual proposal / plan is drawn up and given to the Delegates to discuss and vote upon. IMO, nothing less than a Delegate motion could make the USCF act on this idea.
-
IMO the USCF is proactive, but different folks rank the initiatives differently. Not a bad thing, but can be frustrating on an individual level.
-
Agreed, with a caveat. Even a break-even project can have enormous emotional and organizational costs for a nonprofit.
Quite right. Thanks for the reminder. That was my plan at the beginning, just to bring up the idea. Then when the opinions came I became reflexively very attached to my own trial balloon. I hope, despite that, that my subsequent posts were not too irrational.
Speaking of “small” IT enhancements, over in the Supernationals thread http://main.uschess.org/forums/viewtopic.php?f=5&t=9624 someone was asking for pairings via text-message. Only 140 characters, how hard could it be?
This sounds like something worth looking into.
I don’t know how easy it would be to extract the data from WinTD, but even then there are several questions that come to mind:
-
What does it cost to send text messages? (Probably nothing, I know it’s possible for us to send email messages to cell phones from the USCF’s computers in TN.)
-
How do we collect and update the addresses to send them to?
-
Will multiple text messaging interfaces be needed to send to the various devices players are likely to have? (Does that need to be part of the information collected for #2?)
-
How long would it take to send (up to) 5000 messages? (If it takes 20 minutes or longer to send all the pairings out and have them received, is this still worth considering?)
-
If a player forgets to turn his blackberry (or whatever) off after receiving his pairing on it, is he still penalized if it goes off during the round?
-
Would there be any obstacles to doing this through the hotel’s net connection?
The organizer of a scholastic tournament I am directing in three weeks is planning on texting the pairings to the players. While the process won’t be as seamless as I like it should be fairly efficient. The pairing room staff will upload the pairings from WinTD to custom designed software. The players who have opted in to receive the pairings via text will then all be sent text simultaneously.
I see several advantages to having the pairings sent to cell phones. The players will be less likely to end up at the wrong board or with the wrong color. There will also be less crowding at the pairing sheets.
Grant, you’ve been in the back room at big tournaments, including SuperNationals III. After your upcoming tournament, consider what happens when you try to extrapolate it to a 5000 player tournament with something like 15 sections that are being paired on at least a half dozen computers.
If we pull it off in three weeks Supernationals would be a piece of cake. We have three computers doing 17 sections. Each round is G/30 with pairings going up as soon as the round is over. SuperNationals would be six computers with longer time controls and at least an hour between rounds.
One of the key advantages I see for our event is the speed of getting the players to their games. This isn’t as critical at nationals. I believe one of the TD’s doing pairings for me will also be in the back room at SuperNationals, so I will ask him for his evaluation.
Reducing congestion at the pairing sheets is at least as significant as the time factor. I’ve worried about someone getting hurt in those stampedes at events far smaller than SuperNationals.
And if players start to rely on the texting system always working, what happens when a message gets lost in transit and the player misses getting his pairing? (As I recall, when we were at the Hyatt Regency in Atlanta for the Youth Action, cell phones didn’t work on the level the playing halls were on.)
I don’t see 140 characters as a major limit, though it could get tight.
That’s 122 - 124 characters, depending on if you have to count the line termination.
If the names were a lot longer, it could get close to 140 characters. Including the IDs for both players seems necessary to me if the message could be going to the coach, there could be two players with similar names on the team, and including the rating of the opponent is always helpful, but the round time could be dropped.
The biggest issue I see, aside from collecting the contact information in the first place (imagine 1000 players in line at Chess Control on Friday morning all waiting to give someone their cell phone number) is how to coordinate pairings done on multiple computers with the database of contact information, especially when the contact information needs to be corrected mid-event (which is a virtual certainty, typos happen.)
Maybe this isn’t much different from Phil collecting the pairings from the computers running the various sections to combine them and post them to the website. I’ll ask him next time I talk to him.
I suppose we would also have to consider allowing multiple contact numbers for each player, such as one for the player, one for the player’s parents and one for the player’s coach. (And if one of those numbers just happens to be Grandma’s number back in Poughkeepsie, who’s to know, right?)
It might be worth bringing this idea up with Tom Doan, this might be something that should be incorporated into WinTD.
Oh, I just had a thought. What happens when a round needs to be repaired (in part or in whole) after the text messages go out?
Nothing at this time, AFAIK.
On the tournament entry form, collect cell number and provider.
One interface – SMTP.
http://www.tech-recipes.com/rx/939/sms_email_cingular_nextel_sprint_tmobile_verizon_virgin/
I thought it was 140 bytes max, but Verizon has a 160 byte limit.
TD to player =
5000 messages * 160 bytes/message * 1s/56kB (14 seconds)
- error correction
- hops to the phone
Not if Grant is the TD.
John Hillery thinks yes.
If the pairings are going to be late, might want to send a text notification of the new pairing time, otherwise all the cell-phone users will crush the pairing charts anyway.
If a player signed up for text notification they should still be responsible for checking the pairing chart. In other words, not getting a text message is no excuse for being late.
Try this
That’s 51 characters (name/rating truncated at 20). The character limit has to include the message header, so you need to keep it very short.
Not an issue with an online pairing program. 
One number only! They can use message forwarding if they wish.
Depends of the hotel and the location of the recipient. As long as the pairings are posted normally, and it’s made clear that it’s up to the player to be there on time, I don’t have a problem with it. Whether it’s technically feasible is a different question.
Same thing that happens today. Lots of eye rolls, shoulder shrugs and milling about.
You forgot to factor in the SMTP handshaking and other delays, all of which mean it will take CONSIDERABLY longer than 14 seconds to send out 5000 emails, especially if all you have to work with is a T-1 speed connection (or a fractional portion of a T-1, which is what you usually get at a hotel.)
Phil and Al Losoff are going to kick this around a bit during downtime in the TD room. Phil thinks it might be necessary to bring in a 3rd party service because of anti-spam considerations, especially if trying to use the hotel’s network. That adds costs and may involve other complications.
I didn’t mean to imply it would take only 14 seconds. I just meant it’s not a lot of bytes to send (and 56 kBps is a nice conservative rate). I thought that was the whole point of SMS, doesn’t each message fit in one packet?
But you’re right that I forgot the handshaking, if it’s required for each message that’s a killer. Even at 1/2-second per, that’s over 40 minutes.
I am far from an expert on these things. Your anti-spam comment started the wheels turning. I recall Verizon at one point was going to charge the sender per message, at the time I wondered how that could work? But I didn’t look into it. If a bulk sender has to use a special service to get around some spam filter, or even just to avoid repeated handshaking, then that would explain how Verizon could collect.
Since each message is different, each one would require at least some handshaking, though not necessarily the full SMTP protocol since a connection can send multiple messages.
Some mailer transfer agents limit the total number of messages per connection or insert some delay between messages after the first few, specifically to try to limit large-scale mailings from a single SMTP connection.
There are bulk mail SMTP servers that can bypass most of the SMTP delays, but that’s just to get them into the first mail queue, after which they follow normal protocols to get to their destinations.
Many spammers have taken to using networks of ‘borrowed’ computers to send out their garbage, because they can send many more messages from 100,000 computers under their control. One spammer told the Wall Street Journal that all he needed was about two positive responses for each TEN MILLION emails to make a profit.
In fact, I have my suspicion that the ‘conficker worm’ virus that the anti-virus folks were so on edge about ahead of April 1st may have wound up being a spam relay, because the volume of spam I’ve seen in the last two days has increased several fold, and they’re coming from IP addresses I’ve never gotten messages from before.
Latency could also be an issue if a solution were designed to email all team-related pairings to a coach (who presumably has a computer in the team room).
I look forward to hearing from Mr. Perks whether the text messaging relieves any of the congestion at the pairing sheets.
The most obvious solution to me is not texting the pairings.
It really is not that hard to put the pairings on a web page. Once the pairings are done, posting that as a page on a web site would take less than 5 minutes.
Once on the web page, people can then go to the page to get their pairings. This can be accessed by cell phones with internet capabilities (3G), laptops, IPod Touches, Palm Pilots and the rest that can access WiFi or use an air card.