#include <bits/stdc++.h>
using namespace std;
#define rep(i, a, n) for(int i = a; i < n; i++)
#define rrep(i, a, n) for(int i = a; i >= n; i--)
#define ll long long
#define pii pair<int, int>
#define pll pair<ll, ll>
constexpr ll MOD = 1000000007;
constexpr int IINF = 1001001001;
constexpr ll INF = 1LL<<60;

template<class t,class u> void chmax(t&a,u b){if(a<b)a=b;}
template<class t,class u> void chmin(t&a,u b){if(b<a)a=b;}

// 問題
// https://yukicoder.me/problems/no/5020

const int TIME_LIMIT = 990;
constexpr ll base = 5e17;
const double start_temp = 10000, end_temp = 1000;



int main(){
    auto start_time = chrono::duration_cast<chrono::milliseconds>(chrono::system_clock::now().time_since_epoch()).count();
    random_device seed_gen;
    mt19937 engine(seed_gen());
    int n; cin >> n;
    vector<tuple<ll, ll, ll>> card(n), init(n);
    rep(i, 0, n){
        ll a, b; cin >> a >> b;
        card[i] = {a, b, i};
        init[i] = {a, b, i};
    }
    sort(card.begin(), card.end());
    vector<pair<ll, ll>> ope;
    rep(_, 0, 2){
        rep(__, 0, 25){
            // 1~25回目は表、26~50回目は裏が最大, 最小となるカードについて操作を行う
            auto [a, c, i] = card[0];
            auto [b, d, j] = card[n-1];
            card[0] = {(a+b)/2, (c+d)/2, i};
            card[n-1] = {(a+b)/2, (c+d)/2, j};
            ope.push_back({i, j});
            sort(card.begin(), card.end());
        }
        rep(i, 0, n){
            auto [a, c, j] = card[i];
            card[i] = {c, a, j};
        }
    }

    // 1番のカード
    auto get_card1 = [&](vector<tuple<ll, ll, ll>> card_) ->  pair<ll, ll> {
        rep(i, 0, n){
            auto [a, c, j] = card_[i];
            if(!j){
                return {a, c};
            }
        }
        return {0, 0};
    };

    // スコア
    auto score = [&](pair<ll, ll> card_) -> ll {
        return (ll)(2e6-1e5*log10(max(abs(card_.first-base), abs(card_.second-base))+1));
    };

    // 操作
    auto simulation = [&](vector<pair<ll, ll>> ope_) -> ll {
        vector<tuple<ll, ll, ll>> card_ = init;
        for(auto [u, v]: ope_){
            auto [a, c, i] = card_[u];
            auto [b, d, j] = card_[v];
            card_[u] = {(a+b)/2, (c+d)/2, i};
            card_[v] = {(a+b)/2, (c+d)/2, j};
        }
        return score(get_card1(card_));
    };

    ll best_score = simulation(ope);
    constexpr int ope_size = 50;

    // 山登り
    int cnt = 0, tmp = 0;
    int f1 = 0, f2 = 0;
    while(true){
        auto now_time = chrono::duration_cast<chrono::milliseconds>(chrono::system_clock::now().time_since_epoch()).count();
        if(now_time-start_time >= TIME_LIMIT) break;
        vector<pair<ll, ll>> new_ope = ope;
        if(f1^f2){
            int num = engine()%ope_size;
            int u = 0, v = engine()%(n-1)+1;
            new_ope[num] = {u, v};
        }else if(f1&f2){
            int num = engine()%ope_size;
            int u = engine()%(n-1)+1, v = engine()%(n-1)+1;
            while(u == v) v = engine()%(n-1)+1;
            new_ope[num] = {u, v};
        }else{
            int u = engine()%ope_size, v = engine()%ope_size;
            while(u == v) v = engine()%ope_size;
            swap(new_ope[u], new_ope[v]);
        }
        ll new_score = simulation(new_ope);
        // 温度関数
        double temp = start_temp+(end_temp-start_temp)*((double)(now_time-start_time))/(double)TIME_LIMIT;
        // 遷移確率関数(最大化の場合)
        double prob = min(1.0, exp((double)(new_score-best_score-1)/temp));
        // if(cnt%100000 == 0) cout << new_score-best_score << ' ' << temp << ' ' << prob << endl;
        // 確率probで遷移する
        if (prob > (engine()%IINF)/(double)IINF) { 
            tmp = cnt;
            swap(ope, new_ope);
            swap(best_score, new_score);
        }
        cnt++;
        f2 ^= f1;
        f1 ^= 1;
    }
    //cout << simulation(ope) << ' ' << cnt << endl;
    // 出力
    cout << ope.size() << endl;
    for(auto [u, v]: ope){
        cout << u+1 << ' ' << v+1 << endl;
    }
    return 0;
}