Quote:
Originally Posted by Eugene Church
tysok told me that conditions are already available that may enable us to do this.
I will check it out and let you know.
|
I eagerly await.
On a related, and perhaps more important, matter, the same situation holds true for news stories related to batters (4 for 4), and pitchers (shutouts). I think the way to go is to identify the
story elements and (initially) have 5 or 6 paragraphs (
news items) specific to each. Then the evaluation algorithm could build the story around the situation.
For example (on a batter going 4 for 4) the
story elements might be:
1. Player's recent history (current season)
2. Team's recent history (current season)
3. Summary of batter's hits (singles, doubles, etc.)
4. Effect of batter's performance in the game
5. Batter's description of his performance
6. Summary of which Pitchers he batted against
7. Summary of team's results vs this opponent
8. Team's current place in the standings (and future prospects)
The
story elements for a pitcher pitching a shutout might be:
1. Player's recent history (current season)
2. Team's recent history (current season)
3. Effect of pitcher's performance in the game
4. Summary of which pitches (curve, fastball, etc.) were effective
5. Key innings & trouble spots
6. Pitcher's description of his performance
7. Summary of team's results vs this opponent
8. Team's current place in the standings (and future prospects)
I'll suggest that the combination of applicable
story elements in a news story far outweighs random detail which has no direct relationship to the specific game being described. Even if each
story element only had 3
news items (I recommend 5-6 initially), the combination of 6-8 applicable story elements would give a story all the variety that it needed (i.e., say each news story would have 8
story elements and select 1 of the 3
news items for each). The more
story elements you include - the less news items for each story you need.
Ideally your
news items for each
story element might go something like this...
Items - Description
1 - Excellent
2 - Above Average
3 - Normal
2 - Below Average
1 - Poor
This would result in 9 possible news items (in a bell-shaped curve) for each story element. I'm not a math wiz but I believe that the permutations of 8 story elements times 9 news items would be incredible! You probably never see the same story in a single season twice.