Whew! So this is the start of a huge project that I've been working on. It was initially started to test the oft-repeated claim that "Spike took over the show". However, to do so, it was necessary to establish the dynamics of character screentime before Spike's arrival. This meant that the project expanded and became something, well, a little more interesting.
Now it's a series-wide count of line and scene counts to determine whose show this really is. :)
The juicy part will be when I finally have all seven seasons finished. For now, though, let's look at S1. I'm gonna show the graphs and then provide the raw data so fellow numbers geeks can ooh and ahh over them. First, though!
Methodology:
The method used for this project came about as a result of the intended goal: to compare characters' screentime and line count to see what trends arise throughout the seasons.
The characters involved include all characters who ever had a spot in the opening credit plus Faith. This means that characters who guest star before or after their regular stint still have their lines counted.
I used the transcripts and shooting scripts available on BuffyWorld.
For the line counts:
I first manually went through each transcript, tallying down both the number of lines, total, and the number of lines for each characters. Given that relative numbers are more important than actual numbers, I needed to be able to figure out what percentage of lines each character had in an episode. For this, I needed the total number of lines, even those spoken by incidental characters. The percentage is then simply calculated relative to the amount of lines in the episode (This includes lines by minor characters who are not under assessment).
All numbers were double-checked by opening the transcript in Notepad++ and doing an automated count of the number of times "CHARACTERNAME:" shows up. The manual read-through was necessary to catch any transcription errors or off formatting, while the automated check, obviously, checked my numbers.
For the scene counts:
Counting scenes was slightly more complicated. I ended up using the shooting script when possible. Each new INT or EXT shot counted as a new scene, except establishing shots (stock footage). Also, if the screenwriter labeled a scene as "CONTINUOUS", I didn't count it as a new scene. For scenes that are interrupted by an act break, I counted it as continuous so long as there's no significant time gap. Given that the important factor was figuring out a character's number of scenes relative to each episode (as opposed to comparing total number of scenes on an episode-by-episode basis), I decided that this method, while not perfect, would give satisfactory data.
In some cases in S1, a final shooting script was not available (only a writer's draft). In those cases, I attempted to use the transcript to gather the same information.
In all cases, I made sure to check with the transcript to ensure that I wasn't counting any deleted scenes (or missing any scenes that had been added in). At times, it was necessary to check screenshots to see whether a character was shown in a scene or not.
For the references:
I was able to keep track of the references while doing the line counts. The idea behind this count is to see which characters are often discussed by other characters. This would suggest that the character under discussion is of importance to the plot and to the show.
For a discussion to count, it had to take place while the subject character is not in the scene (or the characters discussing him/her are not aware of his/her presence). My general rule of thumb is that there needed to be three consecutive lines back-and-forth about the other character. Discussions among non-major characters also counted. If a discussion of a character stops and then begins again after a topic change, it's counted as two references. This is the only count that consisted of raw numbers rather than percentages, for obvious reasons.
Though I did track references for S1, I'm waiting until all seasons are completed to crunch those numbers and graph them because this aspect is more relevant in a comparative sense.
Drawbacks:
Using transcripts necessarily means being dependent on the transcribers' formatting. They may divide up a character's line by inserting an action descriptor, for instance. This would be counted as two lines for the character rather than one. However, since I'm looking at percentages relative to counts from the same monologue, it is hoped that internal consistency within the transcript will assuage this.
Also, there's no way to differentiate between short lines and long monologues. As such, this should only serve as a rough guide, and further counts with different methodologies (especially ones incorporating actual time) will be useful.
Here they are, exactly as I tallied them:
1.01:
Scenes: 21
Cordelia: 5 24%
Xander: 4 19%
Willow: 6 29%
Buffy: 14 67%
Angel: 1 5%
Giles: 3 14%
Lines: 425
Buffy: 145 34%
Xander: 36 8%
Willow: 44 10%
Cordelia: 32 8%
Giles: 50 12%
Angel: 10 2%
References:
Buffy: 2
Giles: 1
Cordelia: 1
Xander: 1
1.02:
Scenes 2: 22
Cordelia: 8 36%
Buffy: 17 77%
Willow: 10 45%
Xander: 14 63%
Angel: 2 9%
Giles: 8 36%
Lines: 393
Buffy: 100 25%
Willow: 47 12%
Xander: 70 18%
Giles: 45 11%
Angel: 12 3%
Cordelia: 24 6%
References:
Buffy: 2
1.03:
Scenes: 34
Cordelia: 10 29%
Xander: 13 38%
Willow: 14 41%
Buffy: 26 76%
Giles: 9 26%
Lines: 432
Giles: 60 14%
Buffy: 119 28%
Willow: 57 13%
Xander: 58 13%
Cordelia: 13 3%
References:
Buffy: 1
1.04:
Scenes: 31
Xander: 18 58%
Buffy: 23 74%
Giles: 12 39%
Willow: 15 48%
Cordelia: 2 6%
Angel: 2 6%
Lines: 445
Xander: 93 21%
Buffy: 135 30%
Angel: 12 3%
Giles: 56 13%
Willow: 36 8%
Cordelia: 5 1%
References:
Angel: 1
Xander: 1
1.05:
Scenes: 29
Cordelia: 4 14%
Willow: 14 48%
Xander: 14 48%
Buffy: 21 72%
Giles: 13 45%
Angel: 1 3%
Lines: 464
Buffy: 167 36%
Giles: 83 18%
Willow: 40 9%
Xander: 40 9%
Cordelia: 13 3%
Angel: 9 2%
References:
Cordelia: 1
Buffy: 2
Giles: 1
1.06:
Scenes: 28
Buffy: 18 64%
Willow: 14 50%
Xander: 17 61%
Giles: 8 29%
Lines: 393
Buffy: 101 26%
Willow: 79 20%
Xander: 62 16%
Giles: 43 11%
References:
Xander: 4
Angel: 1
Buffy: 1
1.07:
Scenes: 34
Cordelia: 2 6%
Willow: 12 35%
Xander: 11 32%
Giles: 11 32%
Buffy: 22 65%
Angel: 14 41%
Lines: 367
Buffy: 110 30%
Angel: 54 15%
Willow: 37 10%
Giles: 31 8%
Xander: 34 9%
Cordelia: 6 2%
References:
Angel: 8
Xander: 1
Buffy: 3
1.08:
Scenes: 30
Xander: 10 33%
Willow: 11 37%
Giles: 10 33%
Buffy: 20 67%
Lines: 439
Buffy: 123 28%
Giles: 78 18%
Willow: 59 13%
Xander: 60 14%
References:
Willow: 2
Buffy: 1
1.09:
Scenes: 30
Cordelia: 4 13%
Willow: 12 40%
Xander: 12 40%
Giles: 12 40%
Buffy: 19 63%
Lines: 439
Cordelia: 15 3%
Giles: 70 16%
Buffy: 118 27%
Xander: 51 12%
Willow: 43 10%
References:
Buffy: 1
1.10:
Scenes: 31
Cordelia: 6 19%
Xander: 12 39%
Willow: 13 42%
Buffy: 20 65%
Giles: 7 23%
Lines: 427
Buffy: 128 30%
Willow: 52 12%
Cordelia: 12 3%
Xander: 58 14%
Giles: 52 12%
References:
Cordelia: 1
Buffy: 1
1.11:
Scenes: 38
Cordelia: 20 53%
Xander: 17 45%
Willow: 17 45%
Buffy: 21 55%
Giles: 13 34%
Angel: 2 5%
Lines: 432
Cordelia: 91 21%
Willow: 39 9%
Buffy: 88 20%
Giles: 46 11%
Angel: 14 3%
Xander: 50 12%
References:
Cordelia: 2
1.12:
Scenes: 34
Cordelia: 6 18%
Xander: 11 32%
Willow: 13 38%
Giles: 10 29%
Angel: 6 18%
Buffy: 18 53%
Lines: 404
Xander: 66 16%
Willow: 52 13%
Buffy: 89 22%
Giles: 61 15%
Angel: 26 6%
Cordelia: 26 6%
References:
Buffy: 3
Clicking on the pictures will take you to a larger version.


- For kicks, I took a look to see which characters talked more in the scenes they had. This was done by simply dividing a character's total lines spoken by the number of scenes they had been in. Results as follows:
Buffy: 1423 lines in 239 scenes. 5.95 lines per scene.
Xander: 678 lines in 153 scenes. 4.43 lines per scene.
Willow: 585 lines in 151 scenes. 3.87 lines per scene.
Giles: 675 lines in 116 scenes. 5.82 lines per scene.
Cordelia: 237 lines in 68 scenes. 3.49 lines per scene.
Angel: 137 lines in 28 scenes. 4.89 lines per scene.
What does this mean? Well, it means that Buffy and Giles talk the most when they're actually in scenes. Cordelia and Willow talk the least. And, surprisingly, Angel talks more than Xander when he's actually on screen. Fun, right?
So what's next? I'm well past the halfway point in S2, and I'm also moving forward with the other seasons. I'll post individual overviews of each season, and then there'll be a final post that compares the seasons to note different trends.
Now it's a series-wide count of line and scene counts to determine whose show this really is. :)
The juicy part will be when I finally have all seven seasons finished. For now, though, let's look at S1. I'm gonna show the graphs and then provide the raw data so fellow numbers geeks can ooh and ahh over them. First, though!
Methodology:
The method used for this project came about as a result of the intended goal: to compare characters' screentime and line count to see what trends arise throughout the seasons.
The characters involved include all characters who ever had a spot in the opening credit plus Faith. This means that characters who guest star before or after their regular stint still have their lines counted.
I used the transcripts and shooting scripts available on BuffyWorld.
For the line counts:
I first manually went through each transcript, tallying down both the number of lines, total, and the number of lines for each characters. Given that relative numbers are more important than actual numbers, I needed to be able to figure out what percentage of lines each character had in an episode. For this, I needed the total number of lines, even those spoken by incidental characters. The percentage is then simply calculated relative to the amount of lines in the episode (This includes lines by minor characters who are not under assessment).
All numbers were double-checked by opening the transcript in Notepad++ and doing an automated count of the number of times "CHARACTERNAME:" shows up. The manual read-through was necessary to catch any transcription errors or off formatting, while the automated check, obviously, checked my numbers.
For the scene counts:
Counting scenes was slightly more complicated. I ended up using the shooting script when possible. Each new INT or EXT shot counted as a new scene, except establishing shots (stock footage). Also, if the screenwriter labeled a scene as "CONTINUOUS", I didn't count it as a new scene. For scenes that are interrupted by an act break, I counted it as continuous so long as there's no significant time gap. Given that the important factor was figuring out a character's number of scenes relative to each episode (as opposed to comparing total number of scenes on an episode-by-episode basis), I decided that this method, while not perfect, would give satisfactory data.
In some cases in S1, a final shooting script was not available (only a writer's draft). In those cases, I attempted to use the transcript to gather the same information.
In all cases, I made sure to check with the transcript to ensure that I wasn't counting any deleted scenes (or missing any scenes that had been added in). At times, it was necessary to check screenshots to see whether a character was shown in a scene or not.
For the references:
I was able to keep track of the references while doing the line counts. The idea behind this count is to see which characters are often discussed by other characters. This would suggest that the character under discussion is of importance to the plot and to the show.
For a discussion to count, it had to take place while the subject character is not in the scene (or the characters discussing him/her are not aware of his/her presence). My general rule of thumb is that there needed to be three consecutive lines back-and-forth about the other character. Discussions among non-major characters also counted. If a discussion of a character stops and then begins again after a topic change, it's counted as two references. This is the only count that consisted of raw numbers rather than percentages, for obvious reasons.
Though I did track references for S1, I'm waiting until all seasons are completed to crunch those numbers and graph them because this aspect is more relevant in a comparative sense.
Drawbacks:
Using transcripts necessarily means being dependent on the transcribers' formatting. They may divide up a character's line by inserting an action descriptor, for instance. This would be counted as two lines for the character rather than one. However, since I'm looking at percentages relative to counts from the same monologue, it is hoped that internal consistency within the transcript will assuage this.
Also, there's no way to differentiate between short lines and long monologues. As such, this should only serve as a rough guide, and further counts with different methodologies (especially ones incorporating actual time) will be useful.
Here they are, exactly as I tallied them:
1.01:
Scenes: 21
Cordelia: 5 24%
Xander: 4 19%
Willow: 6 29%
Buffy: 14 67%
Angel: 1 5%
Giles: 3 14%
Lines: 425
Buffy: 145 34%
Xander: 36 8%
Willow: 44 10%
Cordelia: 32 8%
Giles: 50 12%
Angel: 10 2%
References:
Buffy: 2
Giles: 1
Cordelia: 1
Xander: 1
1.02:
Scenes 2: 22
Cordelia: 8 36%
Buffy: 17 77%
Willow: 10 45%
Xander: 14 63%
Angel: 2 9%
Giles: 8 36%
Lines: 393
Buffy: 100 25%
Willow: 47 12%
Xander: 70 18%
Giles: 45 11%
Angel: 12 3%
Cordelia: 24 6%
References:
Buffy: 2
1.03:
Scenes: 34
Cordelia: 10 29%
Xander: 13 38%
Willow: 14 41%
Buffy: 26 76%
Giles: 9 26%
Lines: 432
Giles: 60 14%
Buffy: 119 28%
Willow: 57 13%
Xander: 58 13%
Cordelia: 13 3%
References:
Buffy: 1
1.04:
Scenes: 31
Xander: 18 58%
Buffy: 23 74%
Giles: 12 39%
Willow: 15 48%
Cordelia: 2 6%
Angel: 2 6%
Lines: 445
Xander: 93 21%
Buffy: 135 30%
Angel: 12 3%
Giles: 56 13%
Willow: 36 8%
Cordelia: 5 1%
References:
Angel: 1
Xander: 1
1.05:
Scenes: 29
Cordelia: 4 14%
Willow: 14 48%
Xander: 14 48%
Buffy: 21 72%
Giles: 13 45%
Angel: 1 3%
Lines: 464
Buffy: 167 36%
Giles: 83 18%
Willow: 40 9%
Xander: 40 9%
Cordelia: 13 3%
Angel: 9 2%
References:
Cordelia: 1
Buffy: 2
Giles: 1
1.06:
Scenes: 28
Buffy: 18 64%
Willow: 14 50%
Xander: 17 61%
Giles: 8 29%
Lines: 393
Buffy: 101 26%
Willow: 79 20%
Xander: 62 16%
Giles: 43 11%
References:
Xander: 4
Angel: 1
Buffy: 1
1.07:
Scenes: 34
Cordelia: 2 6%
Willow: 12 35%
Xander: 11 32%
Giles: 11 32%
Buffy: 22 65%
Angel: 14 41%
Lines: 367
Buffy: 110 30%
Angel: 54 15%
Willow: 37 10%
Giles: 31 8%
Xander: 34 9%
Cordelia: 6 2%
References:
Angel: 8
Xander: 1
Buffy: 3
1.08:
Scenes: 30
Xander: 10 33%
Willow: 11 37%
Giles: 10 33%
Buffy: 20 67%
Lines: 439
Buffy: 123 28%
Giles: 78 18%
Willow: 59 13%
Xander: 60 14%
References:
Willow: 2
Buffy: 1
1.09:
Scenes: 30
Cordelia: 4 13%
Willow: 12 40%
Xander: 12 40%
Giles: 12 40%
Buffy: 19 63%
Lines: 439
Cordelia: 15 3%
Giles: 70 16%
Buffy: 118 27%
Xander: 51 12%
Willow: 43 10%
References:
Buffy: 1
1.10:
Scenes: 31
Cordelia: 6 19%
Xander: 12 39%
Willow: 13 42%
Buffy: 20 65%
Giles: 7 23%
Lines: 427
Buffy: 128 30%
Willow: 52 12%
Cordelia: 12 3%
Xander: 58 14%
Giles: 52 12%
References:
Cordelia: 1
Buffy: 1
1.11:
Scenes: 38
Cordelia: 20 53%
Xander: 17 45%
Willow: 17 45%
Buffy: 21 55%
Giles: 13 34%
Angel: 2 5%
Lines: 432
Cordelia: 91 21%
Willow: 39 9%
Buffy: 88 20%
Giles: 46 11%
Angel: 14 3%
Xander: 50 12%
References:
Cordelia: 2
1.12:
Scenes: 34
Cordelia: 6 18%
Xander: 11 32%
Willow: 13 38%
Giles: 10 29%
Angel: 6 18%
Buffy: 18 53%
Lines: 404
Xander: 66 16%
Willow: 52 13%
Buffy: 89 22%
Giles: 61 15%
Angel: 26 6%
Cordelia: 26 6%
References:
Buffy: 3
Clicking on the pictures will take you to a larger version.


- For kicks, I took a look to see which characters talked more in the scenes they had. This was done by simply dividing a character's total lines spoken by the number of scenes they had been in. Results as follows:
Buffy: 1423 lines in 239 scenes. 5.95 lines per scene.
Xander: 678 lines in 153 scenes. 4.43 lines per scene.
Willow: 585 lines in 151 scenes. 3.87 lines per scene.
Giles: 675 lines in 116 scenes. 5.82 lines per scene.
Cordelia: 237 lines in 68 scenes. 3.49 lines per scene.
Angel: 137 lines in 28 scenes. 4.89 lines per scene.
What does this mean? Well, it means that Buffy and Giles talk the most when they're actually in scenes. Cordelia and Willow talk the least. And, surprisingly, Angel talks more than Xander when he's actually on screen. Fun, right?
So what's next? I'm well past the halfway point in S2, and I'm also moving forward with the other seasons. I'll post individual overviews of each season, and then there'll be a final post that compares the seasons to note different trends.
- Current Mood:
bouncy - Current Music:Kate Voegele - Lift Me Up

Comments
Awesome graphs! But... now I want to see how these graphs correlate with episode titles and who wrote what.
*opens tab to Buffyworld*
The largest deviations from the trend are when Angel and Cordy have special episodes just about them. Neat. Otherwise, they're pretty low in terms of lines per episode.
And yep. I was also surprised to see Cordelia beat Buffy (by one percentage point) in any episode. I was also struck by how far below everybody else Willow was. She's pretty quiet this early on. I can't wait to see how that changes as her character develops. :)
But I'm so glad you did, because this is something I've been wanting for so long you don't even know, but I'm much, much too lazy to do it myself. You spoil me! And graphs! You even gave me graphs!
I have to jet before I can really take a look at this, but just knowing that it exists and that you're doing this makes me wild with love. Yay!
I've been working on this for a while. The other seasons should be done more quickly, though, because of the process in which I'm doing them. Speeds up as it goes along. :)
Seriously, I can't wait till all seven seasons are done. There will be an abundance of graphs then! *bounces*
Though I think Cordelia's line:scene ratio is a bit deceptive. Cordelia would often be in a scene, but would be off with her groupies and wouldn't be the focus. Then she'd come across the Scoobies, shoot off a one-liner, then walk off. That's why her line:scene ratio is so low (on par with Willow's). It's not that she doesn't talk a lot. It's that she doesn't talk to the Scoobies a lot this early on.
< / geek >
This is a thing of beauty. Useful referrable-to-in-future-meta-y beauty.
Although Anya could be a ringer.
Now to see if I am right. ;)
There's also some other interesting trends I'm picking up on, but I want to wait until I can crunch the numbers and graph them before commenting. :)
So my bet is that it comes down to Buffy and Spike in the lead.
I think you're going to very surprised at how low Spike's dialogue and scene ratio is. I've seen a much less ambitious tally of season 7 - he isn't there as much as everyone thinks.
The references though, will be very interesting.
Very much looking forward to more of your stats :)
Very much looking forward to more of your stats :)
:)
Edited at 2010-05-26 06:33 pm (UTC)
There is a website that I think is now defunct that did a line count--HellmouthHigh.co.uk was its address, when it existed. But I think your study is better and more careful; I found several mistakes purely by chance in that one (i.e. they claim Angel had zero lines in "Faith, Hope & Trick," but he appeared in a fantasy sequence --> /geeking), and I think they wrote a computer program to calculate the lines and didn't particularly bother double checking that there weren't problems with the transcript format. Your line counts are a bit different so I'm gonna trust you. :)
What's interesting here is that Angel has few lines but is mentioned fairly often; as with Spike later in the show, Angel is a point of conflict among the characters in addition to his own story. And he's Buffy's Boyfriend (becoming one) so that is an issue in and of itself.
Anyway, how crazy is it that Cordy is the first person to beat Buffy in line count? Or that Willow has more lines than Xander in "The Pack"? Hyena-ness not good for the speaking.
Yep. I think I looked through it on the Wayback Machine. But, as you say, they used a different methodology, so I hope my counts will both reinforce and possibly improve their own. :)
What's interesting here is that Angel has few lines but is mentioned fairly often; as with Spike later in the show, Angel is a point of conflict among the characters in addition to his own story. And he's Buffy's Boyfriend (becoming one) so that is an issue in and of itself.
Hell, I was surprised at how often Angel gets talked about. He's the main topic of conversation for Buffy and Willow most of the time. Looking ahead to the raw numbers I have for S7, Spike does get a large number of mentions, as well (Especially in Sleeper).
Anyway, how crazy is it that Cordy is the first person to beat Buffy in line count?
And one of the few. Buffy is rarely beat in that regard (Off the top of my head, Xander beats her in The Replacement but...well...there's two of him in that episode).
Or that Willow has more lines than Xander in "The Pack"? Hyena-ness not good for the speaking.
lol! Cause he's too busy prowling around and eating pigs. :)
<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3<3
<3
*happysigh(
My reaction exactly. :)
Wow. Just, wow.
Sorry. Got distracted. :)
Very interesting. Will look forward to the others!
My favorite nerdy, graphical representation of something has to be this map of Napoleon's invasion of Russia by Charles Joseph Minard that we looked at in my GIS class. He manages to convey about six different aspects of the endeavor in a single image.
http://strangemaps.wordpress.com/2007/1
I am now determined to find out some way to do this with Buffy.
*loves*
I expect the S2 stats to have even more graphs. *bounces with excitement*
I don't think he stole the show exactly, I just think it went in a direction in which he, Buffy, and Willow got the biggest overall story arcs. I'm not a Spike fan, so sometimes this irritates me, but I'm not a big Giles or Xander fan either, so it probably didn't occur to me that they had too little screentime until I went online and saw people complaining about that. Plus I don't think their roles were ever huge anyway. Although I guess that's what some people don't like. I don't think he took the spotlight away from Buffy though, and Willow also got a bigger role as the story went on, so he's not the only character to grow in importance.
Anyway, glad you're back. I've been mostly avoiding interaction with fandom from the moment I entered it, which was years ago, and the few discussions I've had left me winded and nervous, so I understand your need to do it.
Another is that he "stole scenes" or screentime or lines from other major characters. That is relatively easy to quantify, so that's what this project was initially setting out to test. Though, as noted, it's become generalized since it's inception as I find the different trends that are emerging quite fascinating.
Also, myself being spoiled in that I have a lot of the raw numbers for the later seasons, Spike usually has very few lines and even fewer scenes. His relative importance to the story does increase as his relationship with Buffy develops, but that's no different from what occurred with Angel or Riley (both of whom gradually saw more screentime as their connection to Buffy got stronger). As you say, Willow also gets a bigger role as the story develops (It's surprising how little she speaks as first).
A lot of the accusations are from Xander fans who are upset that Xander had such a non-arc in the later seasons. I can sympathize with that, as the raw numbers do seem to show him having fewer lines as the show goes on. I think that "blaming" Spike for this is misguided as it's more to do with the writers not knowing what else to do with Xander and NB's behind-the-scenes alcohol problems.
In any case, it's become an interesting project, and I can't wait to get to the point where I can do a cross-season analysis of the characters' development over the season seasons. *iz a huge nerd*
It's true that Buffy's love interests do get more attention paid to their story arcs than the other male characters, I think. People have also complained that Riley made Xander's soldier memories needless. I haven't seen such complaints about Angel except in reference to the season eight comics, in which Xander does have some prominence and an important role, but it's one that still allows him to be safely in the background.
What I kept hearing about Xander was that there were interviews with the writers and the cast indicating that they just had no more use for Xander, and offered NB the chance to leave the show in season four or five, I don't remember. Maybe I could search Buffy Forums to check for a link, if they have a search function.
People are actually discussing this in the TWOP Buffy forum now, although it quickly to changed to whether or not Spike was a romantic hero, which changed into whether or not Spike saw himself as a bad guy for Buffy in season six. Most people hate him there, although I few Spike fans and Spike neutrals sometimes show up.
TWOP was notoriously unfriendly (aka openly hostile) towards Spike-fans when the show was airing. I don't know if it's mellowed any since the show ended, but they tended to blame Spike for ruining the show (they also thought the show stopped being good after S3). I avoid it like the plague on general principle since I'm far more fond of the post-S3 seasons.
I'm most fond of the first two seasons myself, possibly because of this intangible charm they have I can't really verbalize. I like seasons three through five well enough and there are parts of the last two seasons I enjoy.
What does this mean? Well, it means that Buffy and Giles talk the most when they're actually in scenes. Cordelia and Willow talk the least. And, surprisingly, Angel talks more than Xander when he's actually on screen. Fun, right?
Yes.